# Nasser Ghanemzadeh > Forward Deployed AI Engineer for founder-led B2B SaaS. Ships one narrow AI workflow into your product in 10 business days. Building Vectig in public with Claude Code. Former CEO/CPO at Nivo (acquired, 200K+ users). Nasser Ghanemzadeh has spent 17+ years building products and companies across AI, SaaS, and fintech. As CEO and later CPO at Nivo he built the product organization and shipped AI-powered applications that grew to 200K+ users and 450% revenue growth before the company was acquired. His current commercial focus is the AI Feature Sprint: a 10-business-day engagement for founder-led B2B SaaS teams of 5 to 50 that ships one narrow AI workflow into their product, with fixed scope, fixed price, and a Day-5 prototype guarantee. A focused 2-hour Discovery Day scoping session assesses whether an AI use case is useful, buildable, and worth a Sprint. His live R&D lab is Vectig, an AI-native cash flow forecasting product for pre-seed-to-Series-A founders, built solo with Claude Code, in public. Every workflow pattern he ships into a client codebase has already shipped in Vectig first. He is the author of two books: Forward Deployed AI Engineering (a working guide to the Forward Deployed AI Engineer role: what it is, where it came from, how it is practiced, and how to do it well; 224 pages, May 2026) and Founder Mode (how great founders stay close to product, people, judgment, and execution as companies scale). He writes a weekly newsletter, Notes from the edge of building, and his work has been covered in Forbes, TechCrunch, Al Jazeera, and HuffPost. --- # Nasser Ghanemzadeh | Forward Deployed AI Engineer for founder-led B2B SaaS URL: https://ghanemzadeh.com/ Forward Deployed AI Engineer for founder-led B2B SaaS ## Nasser Ghanemzadeh I embed for 10 business days and ship the AI feature your roadmap keeps deferring. [Book a Sprint →](https://ghanemzadeh.com/ai-feature-sprint/) [Subscribe to the newsletter](https://ghanemzadeh.com/newsletter/) Building Vectig in public. Former CEO/CPO at Nivo (acquired). Built Nivo to 200K users and an acquisition as CEO/CPO. Work covered in Forbes, TechCrunch, Al Jazeera, and HuffPost. **[Vectig](https://vectig.com)**, my own AI-native SaaS, is the live R&D lab. Every pattern I ship into your codebase has already shipped in mine. ### What I'm doing now One commercial focus, one live R&D lab, two books, one newsletter. The hierarchy matters. **Commercial focus:** [10-business-day AI Feature Sprint](https://ghanemzadeh.com/ai-feature-sprint/) for founder-led B2B SaaS teams. One narrow AI workflow shipped into their product. Fixed scope, fixed price, Day-5 prototype guarantee. **Live R&D lab:** [Vectig](https://vectig.com), AI-native cash flow forecasting for pre-seed-to-Series-A founders. Built solo with Claude Code, in public. Every workflow pattern I sell to a Sprint client has shipped in Vectig first. **Authority and method:** two books, [Forward Deployed AI Engineering](https://ghanemzadeh.com/books/forward-deployed-ai-engineering/) and [Founder Mode](https://ghanemzadeh.com/book/). The first explains the role; for the short version, see [what a Forward Deployed AI Engineer is](https://ghanemzadeh.com/forward-deployed-ai-engineer/). The second explains the operating practice underneath it. **Trust and capture:** [weekly newsletter](https://ghanemzadeh.com/newsletter/) covering AI-native product building, Claude Code workflows, and the unglamorous middle of solo SaaS. **Earlier projects:** [Sengi](https://sengi.co), a financial OS for freelancers, built solo as a learning vehicle for AI-native financial tooling. Paused to focus on Vectig and the Sprint. ### Selected work **Nivo**. CEO, then CPO. Built the product organization and shipped AI-powered applications that grew to 200K+ users and 450% revenue growth. Acquired. **Pangouan**. Founding Head of Product (to 2026). Zero to one: no product, no team, no users. Built it from the ground up. **Finnova**. Co-founder. Coworking Space and Accelerator backing early-stage startups. Got to see the ecosystem from the investor side, which changes how you think about everything. **Iran Startups**. Co-founder of a founder community. Earlier, ran [Techly](https://techly.me/), an English-language publication covering the Iranian startup ecosystem (2014 to 2016), and wrote about The Lean Startup methodology in Persian through the [Business of Software](https://businessofsoftware.ir/) blog when almost no one in Iran was talking about it. ### Speaking & media Featured in [Forbes](https://www.forbes.com/sites/elizabethmacbride/2016/04/30/seven-reasons-iran-is-likely-to-be-an-entrepreneurial-powerhouse/), [TechCrunch](https://techcrunch.com/2014/09/02/the-next-tech-startup-ecosystem-to-emerge-iran/), [HuffPost](https://www.huffpost.com/entry/ibridges-the-iranian-runa_b_7627470), and [Al Jazeera](https://www.aljazeera.com/gallery/2016/12/1/inside-irans-silicon-valley). Host of [The Great CEO Podcast](https://www.youtube.com/@ghanemzadeh). ### Get in touch Email: [ghanemzadeh@gmail.com](mailto:ghanemzadeh@gmail.com). For a Sprint, the application form is on [the AI Feature Sprint page](https://ghanemzadeh.com/ai-feature-sprint/). Newsletter ### Notes from the edge of building One email a week. Building solo with Claude Code, AI product distribution, and the unglamorous middle of solopreneurship. No spam. Unsubscribe in one click. ### Quick answers #### What is Nasser Ghanemzadeh working on? Running a 10-business-day AI Feature Sprint for founder-led B2B SaaS teams (one narrow AI workflow shipped into their product, fixed scope, $5K beta). Building Vectig, an AI-native cash flow forecasting product for founders, as a public R&D lab. Previously CEO/CPO at Nivo (acquired, 200K+ users, 450% revenue growth). Earlier work: Sengi (financial OS for freelancers, paused), Pangouan (Founding Head of Product, language learning), Iran Startups ecosystem. #### Where can I work with him? Through the [AI Feature Sprint](https://ghanemzadeh.com/ai-feature-sprint/), the [newsletter](https://ghanemzadeh.com/newsletter/), or by [email](mailto:ghanemzadeh@gmail.com). ### Elsewhere [GitHub](https://github.com/ghanemzadeh) · [LinkedIn](https://www.linkedin.com/in/ghanemzadeh/) · [Goodreads](https://www.goodreads.com/ghanemzadeh) · [AngelList](https://wellfound.com/u/ghanemzadeh) · [Crunchbase](https://www.crunchbase.com/person/nasser-ghanemzadeh) · [SlideShare](https://www.slideshare.net/Ghanemzadeh/presentations) · [F6S](https://www.f6s.com/member/ghanemzadeh) --- # AI Feature Sprint: 10-Day Forward Deployed AI Engineering for B2B SaaS URL: https://ghanemzadeh.com/ai-feature-sprint/ ## Forward Deployed AI engineering for founder-led B2B SaaS. OpenAI and Anthropic each launched deployment subsidiaries in May 2026 to embed engineers inside Fortune 500 companies. The AI Feature Sprint is the same playbook, sized for founder-led SaaS teams that don't have a $250K procurement process. I scope, build, and ship one narrow AI workflow into staging, production, or a production-ready pull request in 10 business days. Built by **Nasser Ghanemzadeh**, ex-CEO/CPO of Nivo (acquired, 200K+ users) · building Vectig in public · featured in Forbes, TechCrunch, Al Jazeera ⚡ Now booking the next Sprint slot · beta capped at 3 teams [Book a $500 Discovery Day](https://ghanemzadeh.com/discovery-day/) [Apply for the $5,000 Sprint](https://tally.so/r/ja5pD6) Beta pricing: $5,000, capped at 3 teams. After beta: $7,500–$10,000. As covered in Forbes · TechCrunch · Al Jazeera · HuffPost ### Your team knows AI matters. The hard part is shipping the right thing. Most SaaS teams are already experimenting with AI. Someone has tried Claude Code. Someone has built a prototype. Someone has suggested adding a chatbot. But useful AI features do not come from vague experimentation. They need a clear workflow, the right data, a narrow scope, product judgment, and an implementation path the team can actually ship. That is what the AI Feature Sprint is designed to create: one managed AI workflow with a clear use case, implementation path, review process, and handoff. ### What is Forward Deployed AI Engineering? Forward Deployed AI Engineering is a delivery model where an engineer embeds inside a customer's team and ships AI capabilities directly into the customer's product and codebase. Not a dashboard. Not a report. Not a Loom. The term comes from Palantir, where it has been the dominant model for two decades. In May 2026, OpenAI launched a $10 billion deployment subsidiary using exactly this title, and Anthropic launched a $1.5 billion deployment subsidiary the same day. Both are aimed at the Fortune 500. The **AI Feature Sprint is the same playbook, sized for founder-led B2B SaaS teams of 5 to 50 people** who can't run a quarter-million-dollar procurement process. More on the role: [what a Forward Deployed AI Engineer is](https://ghanemzadeh.com/forward-deployed-ai-engineer/), and the full guide in the book, [Forward Deployed AI Engineering: A Working Guide to the Hottest Job in Software](https://ghanemzadeh.com/books/forward-deployed-ai-engineering/). ### Examples of narrow AI workflows - **Investor update assistant.** Turn monthly metrics, asks, risks, and progress into a clear investor update draft. - **Finance/runway scenario explainer.** Help founders understand how burn, hiring, churn, or revenue changes affect runway. - **Customer summary generator.** Turn scattered account notes, usage data, and history into a concise customer summary. - **Support ticket summarizer.** Summarize long support threads and surface the issue, urgency, and suggested next step. - **Report generator.** Turn structured data or notes into a useful weekly, customer, or internal report. - **Onboarding assistant.** Guide new users or customers through the first important workflow in your product. ### A concrete example: Investor Update Assistant Here's a workflow I've built into Vectig, the AI-native startup finance product I'm developing in parallel. The Investor Update Assistant pulls structured monthly metrics (MRR, burn, runway, hires, churn, key wins), takes 3–5 sentences of founder-written context, and generates a clean investor update draft. The founder edits, approves, and sends. Average time from open to send: 8 minutes vs. the typical 90+ minutes of blank-page writing. What you'd get in a Sprint: the workflow scoped to your data sources (Stripe, your spreadsheet, your CRM, whatever you use), a working prompt with output quality rubric, integration into your existing product or admin tool, and the handoff document so your team can maintain and iterate. Two weeks. Live by Day 10. **Measured outcome from the live Vectig implementation:** average draft-to-send time 8 minutes, vs. 90+ minutes for blank-page writing. Output quality rubric: 4-of-5 average across accuracy, clarity, honesty, and ask-coverage dimensions in the first 30 sample runs. ### This is not an AI science project. This is not: - A generic AI strategy workshop - A chatbot brainstorm - A 40-page roadmap - An attempt to make your whole product "AI-native" in two weeks - A replacement for your engineering team - A vague exploration of agents, copilots, and buzzwords - A compliance-heavy enterprise AI transformation The sprint is intentionally narrow: one workflow, one clear user problem, one useful shipped outcome. ### Day 0: The readiness check Before we scope a Sprint, we run a 2-hour readiness check on your use case, data, codebase, and team. Below is what a typical founder-led SaaS team looks like at the start. The Sprint begins on Day 1 with these gaps already mapped. ~/sprint-readiness/check.md $ run readiness_check --target founder-saas **use case** vague **data sources** scattered **codebase access** ready **AI provider setup** configured **team availability** partial **success criteria** undefined → recommended path: 2-hour Discovery Day to scope use case and define success criteria. Output simulated. Real diagnostics are generated live during your Discovery Day. Entry offer ### Discovery Day A 2-hour working session designed to assess whether your AI use case is actually useful, buildable, and worth turning into a sprint. $500 **Includes** - 2-hour working session - Use-case clarification - Data and API readiness check - Workflow selection - Risk assessment - Recommended feature scope - Go / no-go memo - Sprint recommendation if there is a strong fit If you continue to the full Sprint within 14 days, the $500 Discovery Day fee is credited toward the Sprint price. [See Discovery Day details →](https://ghanemzadeh.com/discovery-day/) Main offer ### AI Feature Sprint 10 business days. One narrow AI workflow shipped into staging, production, or a production-ready pull request. $5,000 for the first 3 teams. Later: $7,500 to $10,000 after the beta slots. **What is included** - AI use-case selection - Feature scope - UX flow - Product and technical architecture - Prompt and data design - Claude Code workflow - Hands-on implementation of the selected workflow - Working prototype by Day 5 - Staging deployment, production-ready PR, or handoff by Day 10 - Team handoff and enablement session - Loom walkthrough - Handoff notes - 30-day roadmap - Output quality rubric - Reusable workflow playbook [Apply for the $5,000 Sprint](https://tally.so/r/ja5pD6) ### How the 10-business-day sprint works **Day 1: Scope** Define the workflow, user problem, available data, constraints, and success criteria. **Days 2 to 3: Design** Map the UX flow, architecture, prompt and data structure, and implementation plan. **Days 4 to 5: Prototype** Build the first working version and validate the direction. **Days 6 to 9: Implementation** Refine the workflow, handle edge cases, add error states, improve prompts, prepare for handoff or deployment. **Day 10: Handoff** Staging deployment, production-ready pull request, or implementation handoff, depending on your infrastructure and access. ### How a Sprint gets scoped on Day 1 Every Sprint Brief is written around a single sentence: > *For [user], when [trigger or input] happens, the workflow will [process or action] and produce [output], so that [business value].* **Worked example.** "For the founder, when monthly metrics and 3 to 5 sentences of context are added, the workflow will draft a clean investor update covering metrics, risks, and asks, so that the founder can send consistent updates in 8 minutes instead of 90." If we can't agree on the words that fill those five brackets on Day 1, the Sprint isn't ready to start. ### The output quality rubric Every Sprint ships with a 1 to 5 rubric tuned to its specific dimensions. The scale is fixed; the dimensions change per workflow. | Score | Meaning | | --- | --- | | 1 | Unusable. Output is wrong, incoherent, or unsafe. | | 2 | Needs major rewrite. Shape is right; substance requires substantial editing. | | 3 | Usable with significant edits. Team starts from this draft but invests real time refining. | | 4 | Good with minor edits. Small adjustments only. Usable in roughly the time it takes to read. | | 5 | Ready to use. No edits needed. Ships as-is. | Each workflow's rubric specifies its own dimensions: accuracy, clarity, tone, format, completeness. Each comes with worked examples at every score level. ### What we'll take on as a first Sprint - Low risk Internal-only workflows. Drafts and summaries reviewed by a human before anything leaves the company. Most first Sprints sit here, and they should. - Medium risk Workflows touching customer-facing data (support tickets, CRM notes, sales transcripts, internal financial data) with explicit human review. - High risk Legal, medical, compliance, payments, or anything that takes customer-facing actions autonomously. Declined as a first Sprint. We'll help redefine to a lower-risk scope, or wait until trust and tooling are mature. ### Clear scope. Clear expectations. The sprint timeline begins on the agreed kickoff date, scheduled after contract signing and upfront payment. I take one active sprint at a time. Kickoff dates are scheduled on a first-available basis. Client access, decision-maker availability, and timely feedback are required for the 10-business-day timeline. ### This is for you if… - You run or lead product at a B2B SaaS company - You know AI should be part of your product but do not want vague experimentation - You want one useful AI workflow, not a giant AI transformation - Your team can provide access to relevant product, data, or codebase context - You can make decisions quickly during the sprint - You want product judgment and execution, not just code ### This is not for you if… - You want to "add AI" without a real user workflow - You need a full enterprise AI transformation - Your data is inaccessible or not ready - Your team cannot provide access or feedback during the sprint - You want a general chatbot without a clear product use case - You need guaranteed production deployment regardless of infrastructure constraints ### Why work with me I'm Nasser Ghanemzadeh, a former CEO/CPO and product leader with 17+ years building and scaling technology products. - Exited Nivo (200K+ users, 450% revenue growth, acquired). Press coverage in Forbes, TechCrunch, Al Jazeera, HuffPost. - Founding Head of Product at Pangouan. Co-founder of Iran Startups. Former accelerator lead at Finnova. - Currently building Vectig (AI-native startup finance), shipped daily with Claude Code and MCP. This sprint runs on production-grade AI engineering experience from my own products, not generic AI consulting theory. Based in Turkey, available across European, Middle Eastern, and North American business hours. [LinkedIn](https://www.linkedin.com/in/ghanemzadeh/) · [Email](mailto:ghanemzadeh@gmail.com) · [Vectig](https://vectig.com) ### FAQ #### What is a 10-day AI Feature Sprint? A 10-business-day engagement to ship one narrow AI workflow into a B2B SaaS product. Scoping on Day 1, design on Days 2 to 3, prototype by Day 5, implementation on Days 6 to 9, handoff on Day 10. Fixed scope, fixed price. Deliverables include the workflow code committed to your repo, an output quality rubric, a workflow playbook, a Loom walkthrough, and a 30-day post-Sprint roadmap. #### Who is this for? Founder-led B2B SaaS companies with 5 to 50 employees, post-product-market-fit, with paying customers, real data inside their product, and pressure to ship AI features. Not for indie hackers, solopreneurs, or enterprises with formal procurement. #### What does it cost? $500 for a Discovery Day (credited toward the Sprint if you continue within 14 days). $5,000 for the Sprint at beta pricing (first 3 customers). $7,500 to $10,000 after beta. #### What happens if you can't deliver? If by Day 5 there's no working prototype demonstrating the core workflow, you can stop the Sprint and receive 50% back. No subjective judgment, no fine print. Either the prototype runs or it doesn't. #### Why pay you instead of using your own engineers? Three reasons: speed (10 days, not 6 weeks of internal roadmap displacement), judgment (knowing which workflow to ship first is harder than building it), and a fixed-price/fixed-timeline commitment that internal work can't match. Pulling two engineers off your roadmap for six weeks costs roughly $30K in salary plus the opportunity cost of the slipped roadmap. #### How is this different from an AI consulting firm? AI consulting firms deliver decks and roadmaps. This Sprint delivers a working AI workflow in your codebase by Day 10. The category is "Forward Deployed AI Engineering". The same shape of work OpenAI's Deployment Company and Anthropic's deployment arm sell to Fortune 500 customers, at a different price point. #### What kinds of AI workflows do you build? Investor update assistants. Runway scenario explainers. Customer summary generators. Support ticket summarizers. Report generators. Onboarding assistants. The common shape: narrow input, narrow output, single user, measurable improvement, human-in-the-loop review. #### What if our data isn't ready? Then a Sprint isn't the right starting point. The Discovery Day will say so explicitly and recommend what to do first. About half of Discovery Days end with "this isn't a Sprint yet, here's what to do instead." #### Do you handle compliance, security, GDPR, SOC 2? No. Those are explicitly out of scope. The Sprint can produce a workflow that's ready to deploy through your existing compliance posture, but the compliance review itself is your team's work. #### What stack do you work in? Whatever you already use. The Sprint integrates with your existing product, your existing data sources (Stripe, HubSpot, Linear, Slack, Notion, Intercom, Zendesk, your CRM, your admin), and your existing AI provider (Anthropic, OpenAI, AWS Bedrock, Google Vertex AI). The Sprint is too short to introduce new infrastructure. #### Can you guarantee production deployment? No. Production deployment depends on access, infrastructure, and third-party dependencies that aren't fully under our control. If production release is blocked, the deliverable is a staging deployment or a production-ready pull request your team can merge. #### What happens after Day 10? Three paths. Most teams take the workflow and run with it. The rubric and playbook are designed for self-service maintenance. Some come back for a second Sprint with a different workflow. Some sign an ongoing engagement for governance and iteration. All three are explicit options at handoff. ### Terms used on this page **Forward Deployed Engineer** An engineer who embeds inside a customer's team and ships inside their codebase, rather than delivering an external dashboard or report. Originally Palantir's model; adopted by OpenAI and Anthropic in May 2026. **AI Feature Sprint** This offer. 10 business days, one narrow AI workflow, fixed scope, fixed price. **Workflow** A specific process a user runs with clear inputs and outputs, producing a measurable improvement. The unit of work shipped in a Sprint. Not a "feature." **Sprint Brief** The one-to-two-page document written on Day 1 of a Sprint that defines scope, success criteria, deliverables, and explicit out-of-scope items. Part of the contract. **Output Quality Rubric** A 1 to 5 scoring checklist tuned to a specific workflow, used to evaluate generated outputs consistently after handoff. **MCP** Model Context Protocol, the standard for letting AI models talk to external systems. Many Sprint workflows integrate with customer systems via MCP servers. ### Want to ship one useful AI feature? Start with a Discovery Day if you need help choosing the right use case. Apply for the Sprint if you already know what you want to build and need help getting it shipped. [Book a $500 Discovery Day](https://ghanemzadeh.com/discovery-day/) [Apply for the $5,000 Sprint](https://tally.so/r/ja5pD6) --- # Founder Mode | A book by Nasser Ghanemzadeh URL: https://ghanemzadeh.com/book/ **Brian Chesky just reopened the AI Founder Mode conversation for the AI Era.** This book turns the idea into an operating lens. A practical book on how founders stay close to product, people, judgment, and execution as companies scale. ## Founder Mode *How the Great Companies Are Actually Built* In 2024, Paul Graham named "Founder Mode" after Brian Chesky's talk at a YC alumni dinner. The practice is older than the name. HP, Disney, Apple, Pixar, Amazon, NVIDIA, and Costco built their best work this way. This book traces how, and is honest about the costs. Sign up to be notified when the book ships. The first two chapters are yours to read now. You'll also get the newsletter. One email a week, easy unsubscribe. ### Why this book, now Founder Mode is back in the conversation because Brian Chesky is again explaining how great founders operate differently. But the idea is still easy to misunderstand. Founder Mode is not micromanagement. It is not founder ego. It is calibrated involvement: staying close to the decisions, standards, customers, products, and people that shape the company. 01 · What's inside ### Four parts **Foundations** lays out what Founder Mode actually is, separating the practice from the personality around it. It is not micromanagement, and it is not charisma. It is calibrated involvement. **Practices** is the operating manual. The small group of trusted operators. The meetings the founder still runs personally. The decisions the founder still owns. The handoffs that have to happen anyway. **Anchors** is the case-study core. Eight companies, eight chapters, and the specific moves that made each one work. **Returns, Costs, Coda** is the part most books skip. Founder Mode works, and it costs the founder things. The book is honest about both. 02 · Anchor chapters ### Eight companies, eight chapters #### HP Bill and Dave still walking the lab decades after founding, and the practice that institutionalized it. #### Walt Disney The *Snow White* production and a $1.5M bet on feature animation when nobody believed it would work. #### PayPal The small group, the relentless iteration, and the alumni network it produced. #### Apple Jobs in the Newton era. The products that almost shipped, the ones that did, and what he refused to delegate. #### Pixar Ed Catmull's Brain Trust as a structured, durable form of involvement. #### Amazon Bezos's Day 1 doctrine and the six-page memos that enforced it. #### NVIDIA Jensen Huang at Denny's, then thirty years of the same direct-report meetings. #### Costco The buying-club discipline that survived Sinegal's retirement. 03 · About ### About the author I'm Nasser Ghanemzadeh. 17+ years building products and companies across AI, SaaS, and fintech. Previously CEO/CPO at Nivo (acquired). Currently building Vectig solo with Claude Code, and running a 10-business-day [AI Feature Sprint](https://ghanemzadeh.com/ai-feature-sprint/) for B2B SaaS teams. Also by Nasser: [Forward Deployed AI Engineering: A Working Guide to the Hottest Job in Software](https://ghanemzadeh.com/books/forward-deployed-ai-engineering/). More on the [home page](https://ghanemzadeh.com/). 04 · Start reading ### Read the first two chapters Join the launch list and I'll send you the first two chapters now. You'll also be notified when the Kindle version ships. You'll also get the newsletter. One email a week, easy unsubscribe. This book is independent and is not affiliated with Airbnb, Brian Chesky, Paul Graham, Y Combinator, or Invest Like the Best. --- # Books | Nasser Ghanemzadeh URL: https://ghanemzadeh.com/books/ Books ## Two books, one author Different topics, same operating lens. Founder Mode is about how great companies are actually built. Forward Deployed AI Engineering is about a specific job that has become the default way to ship AI into the real world. Both are written for people who actually do the work. [2026 · New ### Forward Deployed AI Engineering A Working Guide to the Hottest Job in Software 224 pages. First edition, May 2026. On May 4, 2026, OpenAI and Anthropic each launched deployment subsidiaries on the same day, both built around the title Forward Deployed Engineer. This book is for the person who wants to do that work. Five parts: the landscape, the mindset and the method, the craft, becoming an FDE, and the economy and future. Read the first chapters →](https://ghanemzadeh.com/books/forward-deployed-ai-engineering/) [2026 ### Founder Mode How the Great Companies Are Actually Built 254 pages. First edition. In 2024, Paul Graham named "Founder Mode" after Brian Chesky's talk at a YC alumni dinner. The practice is older than the name. HP, Disney, Apple, Pixar, Amazon, NVIDIA, and Costco built their best work this way. This book traces how, and is honest about the costs. Read the first two chapters →](https://ghanemzadeh.com/book/) ### Why two books The two books are about the same thing seen from different angles. Founder Mode is about the people who choose to stay close to the work that matters as their companies scale. Forward Deployed AI Engineering is about the work itself, in a specific form, in a specific moment. Read either one first. The connection becomes obvious by the end. ### About the author Nasser Ghanemzadeh is a [Forward Deployed AI Engineer](https://ghanemzadeh.com/forward-deployed-ai-engineer/) for founder-led B2B SaaS. 17+ years building products and companies across AI, SaaS, and fintech. Former CEO and CPO at Nivo (acquired, 200K+ users, 450% revenue growth), Founding Head of Product at Pangouan, and co-founder of Finnova and the Iran Startups community. Work covered in Forbes, TechCrunch, Al Jazeera, and HuffPost. Today he runs the [AI Feature Sprint](https://ghanemzadeh.com/ai-feature-sprint/) and builds [Vectig](https://vectig.com) in public with Claude Code. More on the [home page](https://ghanemzadeh.com/). --- # Forward Deployed AI Engineering | A Working Guide to the Hottest Job in Software URL: https://ghanemzadeh.com/books/forward-deployed-ai-engineering/ **May 4, 2026: OpenAI and Anthropic each launched deployment subsidiaries on the same day, both using the title Forward Deployed Engineer.** This book is for the person who wants to do that work. A working guide to the role OpenAI and Anthropic just put on the map. ## Forward Deployed AI Engineering *A Working Guide to the Hottest Job in Software* First edition. May 2026. Version 1.1. 224 pages. For eighteen months before May 4, 2026, the Forward Deployed Engineer was already the most-discussed engineering role in the industry. Then the two most valuable private AI companies announced, on the same day, that they were creating new entities to put their engineers inside their customers' offices and ship working AI systems. OpenAI capitalized its vehicle at $10 billion. Anthropic at $1.5 billion. Same title. Same playbook. This book is for the person who wants to do that work. [Order on Kindle](https://www.amazon.com/dp/B0H3KP9CVH/) [Order Paperback](https://www.amazon.com/Forward-Deployed-AI-Engineering-Software/dp/B0H3VKKG38/) Or read the first chapters now. Sign up and you'll be notified the moment the book ships. You'll also get the newsletter. One email a week, easy unsubscribe. 01 · Who this is for ### Three readers The book is written for three people specifically. The chapters are ordered for each of them in turn. - **The early-career engineer.** Three to five years into your career. You have read some of the job postings. You recognize something in the description that fits how you want to work. You want a map. - **The mid-career switcher.** Consultant, solutions architect, product manager, ex-founder. You suspect you already have most of the skills. You want to know what is missing. - **The founder or hiring manager.** Building an FDE function for the first time. Needs a primer before going to market. 02 · What you'll be able to do ### By the last chapter The book is engineered around four working outcomes. Finish the book and a reader should be able to do all four. - Explain the role to a hiring manager in your own words, without using "forward deployed" as a magic incantation. - Describe an end-to-end FDE engagement, from the first customer conversation through the production deployment and the synthesis review that turns the engagement into product. - Execute the first week of a real engagement using the templates in the workbook at the back of the book. - Plan a credible path from your current role to a first offer, with a portfolio you can defend and an interview loop you can prepare for. 03 · Structure ### Five parts **Part I. The Landscape.** What the role is, where it came from, and why the AI moment has made it the default operating model of the industry. **Part II. The Mindset and the Method.** The four traits that distinguish people who succeed at this work. The Echo and Delta split that organizes the work into two halves of one craft. The three-phase engagement that gives the work its rhythm. The gravel-road-to-highway pattern that makes the work compound into product. **Part III. The Craft.** Discovery, evaluation, shipping the first slice, integrating with the enterprise stack, and closing the loop. **Part IV. Becoming an FDE.** The paths in. The portfolio. The interview loop. The first year on the inside. **Part V. Economy and Future.** Pricing, outcomes, contracts. Cargo cults and failure modes. The FDE diaspora the labs are producing whether they want to or not. A closing note for the buyer. 04 · Anchor chapters ### Selected chapters #### The Pilot-to-Production Gap Why 95% of enterprise AI pilots produce zero measurable return, and why the gap isn't closed by buying more tokens. #### A Brief History: From Palantir to the Frontier Labs Two decades of FDE practice at Palantir. How it escaped that niche and became the default model for OpenAI, Anthropic, and the labs that followed. #### Echo and Delta: Two Halves of One Craft Echo is the deployment that's working now. Delta is the change that makes the next one work better. Most engineers are good at one. FDEs do both. #### Scope, Validate, Deliver: The Three-Phase Engagement The rhythm every FDE engagement follows, and the specific moves at each phase that separate the engagements that ship from the ones that drift. #### Evals as Specification In FDE work the eval is the spec. How to write one that customers will sign, engineers can build against, and the model can be measured on. #### The FDE Interview Loop What the loop actually tests. The case study, the writing sample, the customer simulation. How to prepare for each, with worked examples. #### Pricing, Outcomes, and Contracts Day rates, project rates, outcome rates, and equity. When each works. The contractual language that protects both sides. #### Cargo Cults and Failure Modes The five ways FDE engagements fail. Each one is recoverable if you catch it by the second week. 05 · Why now ### A snapshot in motion Most of the canonical literature on the role was written between September 2025 and April 2026. Some of it was written while this book was being drafted. The Anthropic and Blackstone joint venture, the OpenAI Deployment Company, the Google Cloud hires, the Big Four repositionings were news, not history, when the manuscript was working through draft. The book is written to still be useful three years from now, but the snapshot dimension is real, and a reader picking it up in 2029 will need to interpret the contemporary references through whatever has happened by then. One principle is most likely to outlast the specific organizations and titles around it. **The artifact is the truth. The code shipping is the substance. The document is the imitation.** Everything else in the book is an unpacking of that. 06 · About the author ### About the author I'm Nasser Ghanemzadeh. 17+ years building products and companies across AI, SaaS, and fintech. Previously CEO/CPO at Nivo (acquired). Currently building Vectig solo with Claude Code. I'm not a Palantir alumnus. I did not invent the FDE motion and I make no claim to having shaped it. What I have done is seventeen years of customer-side software work with the particular intensity that comes from having most often been the one ultimately accountable for whether the software shipped, worked, and earned its keep. I also run a 10-business-day [AI Feature Sprint](https://ghanemzadeh.com/ai-feature-sprint/) for founder-led B2B SaaS teams: ship one narrow AI workflow into their product. Same playbook, sized for teams that don't have a $250K procurement process. More on the [home page](https://ghanemzadeh.com/). Also by Nasser: [Founder Mode](https://ghanemzadeh.com/book/). 07 · Read it ### Read the first chapters Join the launch list and the first chapters land in your inbox now. The full Kindle and paperback editions are available on Amazon today. You'll also get the newsletter. One email a week, easy unsubscribe. ### FAQ #### Who is this book for? Three readers in particular. Early-career engineers (3 to 5 years in) who recognize something in the FDE job description that fits how they want to work. Mid-career switchers (consultants, solutions architects, PMs, ex-founders) who suspect they already have most of the skills. Founders and hiring managers building an FDE function for the first time. If none of those descriptions fits you, the book has failed at scoping. #### How long is it and how is it organized? 224 pages, organized into five parts: the landscape, the mindset and method, the craft, becoming an FDE, and the economy and future. Plus a workbook and annotated references at the back. #### Why this book instead of the other AI books? Most AI books are about the models. This one is about the practice of deploying them inside someone else's company, with all the customer-side friction that implies. It is also written from inside that work, not from the bench. #### When is it available? First edition, May 2026. The full Kindle and paperback editions are available now on Amazon, and you can subscribe to the launch list above to read the first chapters. #### What formats? Kindle and paperback are both available now on Amazon. You can also read the first chapters by subscribing to the launch list. #### Is this affiliated with Palantir, OpenAI, or Anthropic? No. Independent publication. The book discusses these organizations because they shape the role, but the author is not affiliated with any of them. Views are the author's alone. Independent publication. Not affiliated with Anthropic, OpenAI, Google, Palantir, Y Combinator, Andreessen Horowitz, or any other organization mentioned in this book. Product and company names are trademarks of their respective owners. Views are the author's alone. --- # Discovery Day · 2-Hour Scoping Session for AI Feature Sprints | Nasser Ghanemzadeh URL: https://ghanemzadeh.com/discovery-day/ ## Before we scope a Sprint, we scope the problem. A 2-hour working session with Nasser Ghanemzadeh. We look at your use case, your data, your codebase, and your team. By the end you have either a clean Sprint Brief or an honest "this isn't a Sprint yet, here's what to do first" memo. $500. Credited toward your Sprint if you continue within 14 days. [Book a Discovery Day →](https://cal.com/ghanemzadeh/discovery) Next available slot: visible on the booking page. One Discovery Day per week max. As covered in Forbes · TechCrunch · Al Jazeera · HuffPost ### What you walk away with By the end of the 2 hours, you have a working document with all of the following. Not a deck. Not a follow-up email two days later. The document, on the call, ready to forward to your team. - Use-case clarification, written in a single sentence using the Sprint Brief template. - Data and API readiness assessment for the candidate workflow. - Recommended workflow scope (or a recommended deferment, with reasoning). - A go / no-go memo on whether a Sprint is the right next step. - A Sprint Brief draft if there's a strong fit, including success criteria and out-of-scope items. - Honest answers about what fits in 10 business days and what doesn't. ### Who this is for - B2B SaaS founders or product leaders considering shipping an AI feature. - Teams that have tried building AI features and gotten stuck somewhere between prototype and production. - Teams that know AI should be part of their product but want a second opinion before committing engineering time. - Teams comparing Sprint vendors and wanting to validate fit before signing. ### Who this is not for - Teams looking for a generic AI strategy consultation. - Solo indie hackers without a B2B SaaS product. - Enterprise teams with formal procurement or RFP processes. - Anyone looking for a free 30-minute consultation. This is a paid scoping session, not a sales call. ### How a Discovery Day works The 2 hours are structured but not rigid. Rough shape: **Minutes 0–30: Your context.** Current product, team, AI experiments to date, what's worked and what hasn't, what's pulling the AI conversation forward inside your company. **Minutes 30–75: The candidate workflow.** Inputs, outputs, users, data sources, current manual process, what good output looks like, who reviews it before it ships. **Minutes 75–105: Feasibility.** What fits in a 10-day Sprint and what doesn't. Risk classification (low / medium / high). Integration complexity. Where the Day-5 prototype likely lands. **Minutes 105–120: The memo.** Go / no-go recommendation, Sprint Brief draft (if applicable), scoping language, next-step recommendation. Shared as a Google Doc you keep. About half of Discovery Days end with a Sprint recommendation. The other half end with "here's what to do first, come back when X is in place." Both are valid outcomes. The point of a Discovery Day is to find out which one you are, not to convince you to buy a Sprint. ### The credit toward Sprint If you continue to a full Sprint within 14 days of your Discovery Day, the $500 is credited against the Sprint price. So a Discovery Day → Sprint path costs $5,000 total at beta pricing, not $5,500. If we agree the Sprint isn't the right next step, the $500 covers the working document and the 2 hours. No upsell, no follow-up sequence, no retainer pitch. ### What I need from you before the call A short pre-call form, sent after booking. Three things: - **The workflow you're considering.** One paragraph is enough. Examples I've worked through with other founders: investor update assistant, runway scenario explainer, customer summary generator, support ticket triage, onboarding assistant. - **Your stack and data access.** What's your product built in, where does the relevant data live (Stripe, your DB, HubSpot, Linear, Notion, your CRM, your admin), and what AI provider you already use (Anthropic, OpenAI, Bedrock, Vertex) if any. - **The decision-maker question.** Who needs to approve a Sprint, and are they on the call. If they're not, a Discovery Day is fine, but the Sprint recommendation will be harder to act on. That's it. No data dumps, no NDAs to sign before the call. We sign a mutual NDA on the call itself if either side wants one. ### Why work with me I'm Nasser Ghanemzadeh, a former CEO/CPO and product leader with 17+ years building and scaling technology products. - Exited Nivo (200K+ users, 450% revenue growth, acquired). Press coverage in Forbes, TechCrunch, Al Jazeera, HuffPost. - Founding Head of Product at Pangouan. Co-founder of Iran Startups. Former accelerator lead at Finnova. - Currently building Vectig (AI-native startup finance), shipped daily with Claude Code and MCP. **Speaking on AI-native product development:** OMR Festival, re:publica, GITEX EUROPE. This sprint runs on production-grade AI engineering experience from my own products, not generic AI consulting theory. Based in Turkey, available across European, Middle Eastern, and North American business hours. [LinkedIn](https://www.linkedin.com/in/ghanemzadeh/) · [Email](mailto:ghanemzadeh@gmail.com) · [Vectig](https://vectig.com) ### FAQ #### What if we decide not to do a Sprint after the Discovery Day? Then you walk away with the working document, the go / no-go memo, and 2 hours of focused outside-perspective on your AI roadmap. No follow-up sales sequence. You're free to take the memo to another vendor, to your internal team, or to nobody at all. #### Do you sign an NDA? If either side wants a mutual NDA, we sign one on the call itself before getting into specifics. I don't ask for NDAs to be signed before booking, and I don't expect you to. #### What if our data isn't ready? That's one of the most common Discovery Day outcomes: "the workflow makes sense, but the data foundation isn't there yet." The memo will say exactly what to do first, usually a 2 to 4 week data-readiness project that you can run internally, and we revisit Sprint readiness when it's done. #### How is this different from a free consultation call? A free call is sales discovery. This is a paid working session. The $500 buys two things: focused 2-hour attention with no agenda except your problem, and a written document you keep regardless of outcome. The economics force both sides to take it seriously. #### What if I'm not sure I want a Sprint yet? Then a Discovery Day is exactly the right starting point. About half of them end with a deferment recommendation. That's a feature, not a failure mode. #### Can a co-founder or technical lead join? Yes. Up to three people from your side on the call. Beyond three the conversation slows down. If you want broader team alignment, the right pattern is: founder + tech lead on the Discovery Day, then I record a Loom walkthrough of the memo for everyone else. #### What if our use case is in a high-risk category (legal, medical, compliance, payments)? The Discovery Day still works, but the Sprint recommendation will likely be "not yet" or "redefine scope to a lower-risk adjacent workflow." High-risk first Sprints don't make sense for teams new to AI deployment. The memo will explain what does. #### Do you work with companies outside B2B SaaS? Mostly no. The methodology, examples, and Sprint structure are built around founder-led B2B SaaS at 5 to 50 people, post-product-market-fit. If you're outside that, the Discovery Day may still be useful but the Sprint may not be the right vehicle. ### Two hours, one document, a clear next step. Book a Discovery Day if you want a focused outside perspective on whether the AI feature you're considering is ready to ship in 10 business days. If you'd rather start with the Sprint details, the full breakdown is at [/ai-feature-sprint](https://ghanemzadeh.com/ai-feature-sprint/). [Book a Discovery Day →](https://cal.com/ghanemzadeh/discovery) [Read the Sprint details](https://ghanemzadeh.com/ai-feature-sprint/) --- # Forward Deployed AI Engineer vs Solutions Engineer | Nasser Ghanemzadeh URL: https://ghanemzadeh.com/forward-deployed-ai-engineer-vs-solutions-engineer/ [Forward Deployed AI Engineer](https://ghanemzadeh.com/forward-deployed-ai-engineer/) ## Forward Deployed AI Engineer vs solutions engineer A Forward Deployed AI Engineer and a solutions engineer share the same technical, customer-facing instinct, but they work for different sides and ship different things. The Forward Deployed AI Engineer builds an AI workflow into the customer's own product and owns the result. The solutions engineer sells, demos, and integrates a vendor's product on the vendor's behalf. By [Nasser Ghanemzadeh](https://ghanemzadeh.com/). Last updated June 2026. ### The short answer If the code you ship lives in the customer's repository and the win is the customer's outcome, that is forward deployed work. If the code you ship is a demo or integration that helps close and onboard a customer onto your employer's platform, that is solutions engineering. The two roles look similar from the outside because both put a technical person in front of a customer. The difference is whose product gets built and who you answer to. ### Side by side | Dimension | Forward Deployed AI Engineer | Solutions engineer | | --- | --- | --- | | Works for | The customer's outcome, via a deployment company or a direct engagement | The software vendor selling the product | | Primary deliverable | A production AI workflow in the customer's codebase | Demos, proofs of concept, and integrations | | Where the code lives | The customer's product and repository | The vendor's platform and demo environments | | Measured on | A shipped workflow the customer's team can run and maintain | Pipeline, deals closed, and successful onboarding | | AI focus | Prompt and context design, data integration, evaluation rubrics | Showing how the vendor's AI features fit the customer's stack | ### Where they overlap Both roles need to read a customer's real environment quickly, translate a business problem into a technical plan, and stay credible in a room with engineers and executives at the same time. Both live with ambiguity and incomplete data. The instincts transfer cleanly, which is why solutions engineers are one of the most common backgrounds for people moving into forward deployed AI work. ### How to choose between them Choose by what you want to own. If you want to close deals and help many customers adopt one product, solutions engineering fits. If you want to build the actual feature and own whether it works in production, the Forward Deployed AI Engineer role fits. One more practical signal: forward deployed work rewards depth on one customer at a time, while solutions engineering rewards breadth across a pipeline. Neither is more senior than the other. They optimize for different outcomes. ### Related - [What is a Forward Deployed AI Engineer?](https://ghanemzadeh.com/forward-deployed-ai-engineer/) - [How to become a Forward Deployed AI Engineer](https://ghanemzadeh.com/how-to-become-a-forward-deployed-ai-engineer/) - [The book: Forward Deployed AI Engineering](https://ghanemzadeh.com/books/forward-deployed-ai-engineering/) ### Frequently asked questions #### Can a solutions engineer become a Forward Deployed AI Engineer? Often, yes. Solutions engineers already have the customer-facing instinct and the integration skills. The shift is from selling and configuring a vendor's product to building and owning a production AI workflow inside the customer's codebase, plus the evaluation discipline to prove it works. #### Who does each role work for? A solutions engineer almost always works for the software vendor and is measured on sales and onboarding. A Forward Deployed AI Engineer works on the customer's outcome, whether employed by a deployment company or engaged directly, and is measured on shipping a workflow the customer's team can run. #### Does a solutions engineer write production code? Sometimes, but usually demos, proofs of concept, and integrations rather than features that live permanently in the customer's product. A Forward Deployed AI Engineer's main deliverable is production code in the customer's repository, with edge cases, logging, and a handoff. --- # What Is a Forward Deployed AI Engineer? | Nasser Ghanemzadeh URL: https://ghanemzadeh.com/forward-deployed-ai-engineer/ The role ## What is a Forward Deployed AI Engineer? A Forward Deployed AI Engineer is an engineer who embeds inside a customer's team and ships AI capabilities directly into the customer's product and codebase, rather than handing over a strategy deck, a dashboard, or a demo. The role pairs hands-on AI engineering with the customer-facing judgment to decide what to build. [Read the book →](https://ghanemzadeh.com/books/forward-deployed-ai-engineering/) [See it as a service](https://ghanemzadeh.com/ai-feature-sprint/) By [Nasser Ghanemzadeh](https://ghanemzadeh.com/). Last updated June 2026. ### Where did the term come from? The phrase is borrowed from Palantir, which built its business on engineers who deploy into a customer's organization and deliver outcomes inside that customer's systems, not from a vendor booth. The model spread because it solved a specific failure: software that works in a demo and dies in the customer's real environment. In May 2026 the term went mainstream. OpenAI and Anthropic each launched a deployment subsidiary within a day of each other, both using the title Forward Deployed Engineer, both aimed at getting frontier models into enterprise products rather than selling API access and hoping. Adding AI to the title marks the shift: the capability being deployed is no longer just software, it is an AI workflow. ### What does a Forward Deployed AI Engineer actually do? They take one narrow use case from a vague idea to a working AI feature inside the customer's product, end to end, and they own every step in between. The work is concrete, not advisory. A typical engagement runs through: - **Scoping.** Turn "we should add AI" into one workflow with a named user, a clear input, and a clear output. - **Architecture and UX.** Decide where the workflow lives in the product and how a person actually triggers and reviews it. - **Prompt and context design.** Build the model interaction, including the system prompt, the retrieved context, and the guardrails. - **Data integration.** Connect the workflow to the customer's real data and systems, often through the Model Context Protocol. - **Implementation.** Write production code in the customer's repository, with edge cases, fallbacks, error states, and logging. - **Evaluation.** Ship a written rubric that defines what good output looks like and how the team scores it over time. - **Handoff.** Hand the team code they can maintain, plus the rubric and a short playbook, so the workflow survives without the engineer. The output is code in the customer's repo. Not a dashboard. Not a report. Not a Loom. ### How is a Forward Deployed AI Engineer different from a solutions engineer or ML engineer? All three are technical, but they ship different things to different places. A Forward Deployed AI Engineer ships an AI workflow into the customer's own product and owns the outcome. | Role | What they ship | Where the work lives | Optimizes for | | --- | --- | --- | --- | | Forward Deployed AI Engineer | A working AI workflow built into the customer's product | The customer's codebase and product | A shipped outcome the customer's team can run and maintain | | Solutions engineer | Integrations, demos, and configuration of a vendor's product | The boundary between vendor and customer | Closing and onboarding the customer onto the vendor's platform | | ML engineer | Models, training pipelines, and inference infrastructure | The model and data platform | Model performance, scale, and reliability | | AI consultant | Strategy, roadmaps, and recommendations | Documents and meetings | Advice and direction, with execution left to others | For a side-by-side that goes deeper on the closest comparison, see [Forward Deployed AI Engineer vs solutions engineer](https://ghanemzadeh.com/forward-deployed-ai-engineer-vs-solutions-engineer/). ### What skills does the role require? Less model research, more product engineering and judgment. The hard part is usually deciding what to build and proving it works, not calling the API. - **Product engineering.** Shipping real features into a real codebase, with the discipline that production demands. - **Prompt and context design.** Getting reliable output from frontier models, including retrieval and structured context. - **Data integration.** Connecting models to messy customer systems and APIs, often via [MCP](https://ghanemzadeh.com/glossary/#mcp). - **Evaluation.** Defining what good output is and scoring it, so quality is measured rather than asserted. - **Customer-facing judgment.** Scoping under pressure, saying no to the wrong workflow, and protecting a fixed timeline. ### How do you become a Forward Deployed AI Engineer? Most people who do this well arrive from product engineering, solutions or consulting, or a founder background, then prove they can ship one narrow AI workflow end to end. The fastest credible path is to build and ship a real workflow, write the evaluation rubric for it, and be able to walk someone through the code. For a step-by-step version, see [how to become a Forward Deployed AI Engineer](https://ghanemzadeh.com/how-to-become-a-forward-deployed-ai-engineer/). ### Forward Deployed AI Engineering for founder-led B2B SaaS The enterprise version of this role is built for Fortune 500 procurement. The same work, sized for a 5-to-50-person founder-led SaaS team, is the AI Feature Sprint: one narrow AI workflow shipped into the product in 10 business days, fixed scope and fixed price. That is the work I do. Every pattern I ship into a customer's codebase has already shipped in [Vectig](https://vectig.com), my own AI-native product and live R&D lab. If you want to see the engagement, start with the [AI Feature Sprint](https://ghanemzadeh.com/ai-feature-sprint/) or a [Discovery Day](https://ghanemzadeh.com/discovery-day/). The book ### A working guide to the role [Forward Deployed AI Engineering: A Working Guide to the Hottest Job in Software](https://ghanemzadeh.com/books/forward-deployed-ai-engineering/) is the long-form version of this page: what the role is, where it came from, how it is practiced, and how to do it well. Written from inside the work, in the months after OpenAI and Anthropic put the title on the map. ### About the author I'm Nasser Ghanemzadeh, a Forward Deployed AI Engineer for founder-led B2B SaaS. 17+ years building products and companies across AI, SaaS, and fintech. Previously CEO/CPO at Nivo (acquired, 200K+ users, 450% revenue growth). Work covered in Forbes, TechCrunch, Al Jazeera, and HuffPost. I run the [AI Feature Sprint](https://ghanemzadeh.com/ai-feature-sprint/), build [Vectig](https://vectig.com) in public with Claude Code, and wrote the book on this role. More on the [home page](https://ghanemzadeh.com/). ### Frequently asked questions #### Is a Forward Deployed AI Engineer the same as a forward deployed software engineer? It is the same delivery model applied to AI work. A forward deployed software engineer embeds with a customer and ships software into their systems. A Forward Deployed AI Engineer does the same, but the capability shipped is an AI workflow: prompt design, model integration, data plumbing, and an evaluation rubric, built into the customer's product. #### Do you need a machine learning background to be a Forward Deployed AI Engineer? No. The role is about applying existing models, not training them. The core skills are product engineering, prompt and context design, data integration, evaluation, and customer-facing judgment. A research or MLOps background helps in some engagements but is not the entry requirement. #### Is Forward Deployed AI Engineering only for large enterprises? No. Palantir, OpenAI, and Anthropic run forward deployed teams aimed at Fortune 500 procurement, but the same shape of work fits founder-led B2B SaaS companies of 5 to 50 people. The engagement is just sized down: one narrow workflow, fixed scope, shipped in days rather than quarters. #### What tools does a Forward Deployed AI Engineer use? Frontier models from providers like Anthropic and OpenAI, an agentic coding tool such as Claude Code, the Model Context Protocol (MCP) to connect models to the customer's systems, and a written evaluation rubric to keep output quality measurable. The customer's own stack is the deployment target. #### How is this different from AI consulting? A consultant delivers advice: a strategy, a roadmap, a recommendation. A Forward Deployed AI Engineer delivers working code in the customer's repository. The judgment is packaged with the execution. The deck, if there is one, describes something that already runs in production. Newsletter ### Notes from the edge of building One email a week on Forward Deployed AI Engineering, building AI-native SaaS solo with Claude Code, and the unglamorous middle of shipping. No spam. Unsubscribe in one click. --- # Forward Deployed AI Engineering Glossary | Nasser Ghanemzadeh URL: https://ghanemzadeh.com/glossary/ [Forward Deployed AI Engineer](https://ghanemzadeh.com/forward-deployed-ai-engineer/) ## Forward Deployed AI Engineering glossary The working vocabulary of Forward Deployed AI Engineering for founder-led B2B SaaS. Plain definitions of the role, the offer, the deliverables, and the tools, written the way they are actually used. By [Nasser Ghanemzadeh](https://ghanemzadeh.com/). Last updated June 2026. **Forward Deployed AI Engineer** An engineer who embeds inside a customer's team and ships AI capabilities directly into the customer's product and codebase, rather than delivering a deck, a dashboard, or a demo. The delivery model comes from Palantir. OpenAI and Anthropic both launched deployment subsidiaries using it in May 2026. See the full explainer: [what is a Forward Deployed AI Engineer?](https://ghanemzadeh.com/forward-deployed-ai-engineer/) **AI Feature Sprint** A ten-business-day engagement that ships one narrow AI workflow into a customer's product. Fixed scope, fixed price, end-to-end delivery and handoff. See the [AI Feature Sprint](https://ghanemzadeh.com/ai-feature-sprint/). **Discovery Day** A two-hour paid working session, priced at $500, to assess whether an AI use case is useful, buildable, and worth a Sprint. The output is a go/no-go memo within 24 hours, credited toward a Sprint if the customer continues. See [Discovery Day](https://ghanemzadeh.com/discovery-day/). **Workflow** A specific process a user runs, with a clear input and a clear output, that produces a measurable improvement. It is the unit of work shipped in a Sprint. Not a feature, and not a platform. **Output Quality Rubric** A practical checklist, usually on a 1-to-5 scale, that defines what good output from an AI workflow looks like, how the customer's team reviews it, and how the workflow improves over time. It ships as a deliverable with every Sprint. **Sprint Brief** The one-to-two-page document written on Day 1 of a Sprint and signed off by the customer's decision-maker, defining scope, success criteria, deliverables, and what is explicitly out of scope. It is part of the contract. **Workflow Playbook** A short handoff document covering a workflow's purpose, inputs, expected outputs, review steps, edge cases, and maintenance guidance, so the customer's team can run and improve it without the engineer. **Model Context Protocol (MCP)** The open standard Anthropic introduced for letting AI models talk to external systems and data. Many AI workflows connect to a customer's tools and data through MCP servers. **Claude Code** Anthropic's command-line tool for agentic coding. A primary delivery tool for shipping AI workflows directly into a codebase. **AI-native product** A product built around AI workflows from the start, rather than a traditional product with an AI feature added on later. **Vectig** Nasser Ghanemzadeh's AI-native startup finance product: cash-flow forecasting, runway scenarios, and investor update drafting. It serves as the live R&D lab where workflow patterns are proven before they ship to customers. See [vectig.com](https://vectig.com). ### Related - [What is a Forward Deployed AI Engineer?](https://ghanemzadeh.com/forward-deployed-ai-engineer/) - [How to become a Forward Deployed AI Engineer](https://ghanemzadeh.com/how-to-become-a-forward-deployed-ai-engineer/) - [The book: Forward Deployed AI Engineering](https://ghanemzadeh.com/books/forward-deployed-ai-engineering/) --- # How to Become a Forward Deployed AI Engineer | Nasser Ghanemzadeh URL: https://ghanemzadeh.com/how-to-become-a-forward-deployed-ai-engineer/ [Forward Deployed AI Engineer](https://ghanemzadeh.com/forward-deployed-ai-engineer/) ## How to become a Forward Deployed AI Engineer You become a Forward Deployed AI Engineer by proving you can take one narrow AI use case from a vague idea to a working feature in a real product, then explain every decision behind it. The fastest credible path is to build and ship one workflow end to end, evaluation rubric included, and be able to walk someone through the code. By [Nasser Ghanemzadeh](https://ghanemzadeh.com/). Last updated June 2026. ### Who already has a head start Most people who do this well arrive from one of three backgrounds: product and software engineering, solutions engineering or consulting, or a founder background. Product engineers bring the ability to ship into a real codebase. Solutions engineers and consultants bring the customer-facing instinct. Founders bring judgment about what is worth building. You do not need all three. You need to be honestly strong in one and willing to build the others. ### The path, step by step - **Get genuinely good at shipping software.** The role is product engineering first. If you cannot ship a feature into a real codebase with edge cases and error states, start there. - **Use frontier models every day.** Work with the current Anthropic and OpenAI models until their behavior is intuitive. Adopt an agentic coding tool like Claude Code as your default way of building. - **Learn prompt and context design.** Reliable output is an engineering problem: system prompts, retrieved context, structured outputs, and guardrails. Treat it like code, not like magic words. - **Learn to connect models to real systems.** Understand the [Model Context Protocol (MCP)](https://ghanemzadeh.com/glossary/#mcp) and how to wire a model into a customer's data and tools safely. - **Build one narrow workflow, end to end.** Pick a real use case with a named user, a clear input, and a clear output. Ship it inside a real product or a convincing clone. Narrow beats ambitious. - **Write the evaluation rubric.** Define what good output looks like on a simple scale and score against it. This discipline is what separates an engineer from someone who got one lucky demo. - **Develop customer-facing judgment.** Practice scoping under pressure, saying no to the wrong workflow, and protecting a fixed timeline. The hardest skill is resisting the pull of a vague request. - **Show the work.** Record a short walkthrough of the code, the prompts, and the rubric. Being able to defend every decision out loud is the real credential. ### What to study, and what to skip Study product engineering, prompt and context design, data integration, and evaluation. You can skip the machine learning research track. The role applies existing models; it does not train them. A research or MLOps background is useful in some engagements, but it is not the entry requirement, and chasing it first is the most common way to delay the thing that actually gets you hired: a shipped workflow. Go deeper ### The book [Forward Deployed AI Engineering: A Working Guide to the Hottest Job in Software](https://ghanemzadeh.com/books/forward-deployed-ai-engineering/) covers the full path in depth, including a part on becoming an FDE, plus a workbook. It is written from inside the work. ### Related - [What is a Forward Deployed AI Engineer?](https://ghanemzadeh.com/forward-deployed-ai-engineer/) - [Forward Deployed AI Engineer vs solutions engineer](https://ghanemzadeh.com/forward-deployed-ai-engineer-vs-solutions-engineer/) - [Forward Deployed AI Engineering glossary](https://ghanemzadeh.com/glossary/) ### Frequently asked questions #### What background do most Forward Deployed AI Engineers come from? Three backgrounds are most common: product or software engineers who can ship into a real codebase, solutions engineers and consultants who already work with customers, and ex-founders who have built and sold something. The shared trait is the ability to ship and to make decisions in front of a customer. #### How long does it take to become a Forward Deployed AI Engineer? If you can already ship software, the AI-specific layer (prompt and context design, model integration, evaluation) takes weeks of focused practice, not years. The honest gate is not study time. It is whether you have shipped one real AI workflow end to end and can defend every decision in it. #### What should I build to prove I can do the work? Build one narrow workflow with a named user, a clear input, and a clear output, inside a real product or a convincing clone. Ship it, write an evaluation rubric for it, and record a short walkthrough of the code and prompts. That single artifact is more persuasive than any certificate. --- # Notes from the edge of building | Nasser Ghanemzadeh URL: https://ghanemzadeh.com/newsletter/ 01 · The Newsletter ## Notes from the edge of building Weekly notes on building AI-native SaaS solo. Product, distribution, founder ops, and the unglamorous middle. Free. No spam. Unsubscribe in one click. 02 · Latest ### Latest articles - [July 15, 2026 #### What Exactly Is a Forward Deployed AI Engineer? It is not a sales engineer, a consultant, or a CSM with a GitHub account. Every word in the title is doing work.](https://ghanemzadeh.substack.com/p/what-exactly-is-a-forward-deployed?utm_source=ghanemzadeh.com&utm_medium=newsletter_page) - [July 8, 2026 #### Why 95% of Enterprise AI Pilots Fail (And It's Not the Models) The most important number in enterprise AI is not a benchmark score. It is the gap between pilot and production.](https://ghanemzadeh.substack.com/p/why-95-of-enterprise-ai-pilots-fail?utm_source=ghanemzadeh.com&utm_medium=newsletter_page) - [July 1, 2026 #### Forward Deployed Engineering at $200/Hour The brutal math of fixed-price AI work, and why every operator who underprices this work goes broke profitably.](https://ghanemzadeh.substack.com/p/forward-deployed-engineering-at-200hour?utm_source=ghanemzadeh.com&utm_medium=newsletter_page) - [June 24, 2026 #### What Forward Deployed AI Engineering Actually Is The most important new role in B2B software is the one almost nobody can define correctly. By the end of this article, you will.](https://ghanemzadeh.substack.com/p/what-forward-deployed-ai-engineering?utm_source=ghanemzadeh.com&utm_medium=newsletter_page) - [June 17, 2026 #### The Five Mistakes Every Forward Deployed AI Engagement Makes Operational wisdom from running engagements yourself, watching other operators run theirs, and seeing in-house teams attempt them. These five mistakes show up over and over.](https://ghanemzadeh.substack.com/p/the-five-mistakes-every-forward-deployed?utm_source=ghanemzadeh.com&utm_medium=newsletter_page) - [June 10, 2026 #### Forward Deployed Engineering Without the $250K Floor Why the hottest role in tech needs to exist in two sizes, and why most founder-led B2B SaaS companies are quietly being priced out of the one they actually need.](https://ghanemzadeh.substack.com/p/forward-deployed-engineering-without?utm_source=ghanemzadeh.com&utm_medium=newsletter_page) - [June 5, 2026 #### The Three Risk Tiers of AI Deployment Every AI workflow lives in one of three risk tiers. Most teams don’t know which one they’re shipping until something goes wrong. Here is the taxonomy.](https://ghanemzadeh.substack.com/p/the-three-risk-tiers-of-ai-deployment?utm_source=ghanemzadeh.com&utm_medium=newsletter_page) - [June 3, 2026 #### How a Forward Deployed Engineer Scopes an AI Workflow in 2 Hours Why most internal AI projects fail at scoping, not implementation, and the single 2-hour session that decides whether the next 10 days produce a shipped workflow or a refund conversation.](https://ghanemzadeh.substack.com/p/how-a-forward-deployed-engineer-scopes?utm_source=ghanemzadeh.com&utm_medium=newsletter_page) - [May 29, 2026 #### The Painkiller Test A 30-second test for any AI workflow idea, and why most “AI features” fail it.](https://ghanemzadeh.substack.com/p/the-painkiller-test?utm_source=ghanemzadeh.com&utm_medium=newsletter_page) - [May 27, 2026 #### How a Forward Deployed Engineer Runs a 10-Day AI Sprint Day by day inside a fixed-scope, fixed-price AI engagement, including the day that quietly decides whether the sprint ends well.](https://ghanemzadeh.substack.com/p/how-a-forward-deployed-engineer-runs?utm_source=ghanemzadeh.com&utm_medium=newsletter_page) - [May 25, 2026 #### Narrow Is the Moat Why we won't sell you a bigger package, even if you ask.](https://ghanemzadeh.substack.com/p/narrow-is-the-moat?utm_source=ghanemzadeh.com&utm_medium=newsletter_page) - [May 21, 2026 · Founder Operations #### The Day-4 Trap What we learned about integrity, refunds, and reputation from running fixed-price AI sprints.](https://ghanemzadeh.substack.com/p/the-day-4-trap?utm_source=ghanemzadeh.com&utm_medium=newsletter_page) [All posts on Substack →](https://ghanemzadeh.substack.com/archive?utm_source=ghanemzadeh.com&utm_medium=newsletter_page) 03 · What you'll get ### What you'll get A short, weekly email from someone in the trenches building AI-native SaaS solo with Claude Code. No frameworks borrowed from Twitter threads. No promises of overnight scale. Just what is working, what is not, and what I am trying next. 04 · What I write about ### Five topics - **Forward Deployed AI Engineering in practice.** What works in the field, what the labs are getting right and wrong. - **Building AI-native products solo.** Vectig as the live case study, including the failures, plus lessons from Sengi, an earlier build I paused. - **Distribution without paid ads.** LinkedIn, warm outreach, the referrer network. - **Founder operations.** Cash flow, hiring, scope, the unglamorous middle. - **Claude Code workflows.** The specific patterns that make a one-person shop ship like a small team. 05 · A taste ### A taste of the voice > One narrow workflow. Not a platform, not an AI strategy, not a copilot, not a chatbot the marketing team thought of. One specific use case with a clear input, a clear output, and a person on the customer's team who will actually use it. Examples we'll come back to: an investor update assistant, a customer summary generator, a support ticket summarizer. > The workflow lives inside the customer's existing software. A feature in their app, an internal admin tool, or a production-ready pull request their team can merge. Not a standalone tool we host. Not a demo on a Loom. Code in their repo. End-to-end. > From the Forward Deployed AI Onboarding Handbook. Representative of this newsletter's voice. 06 · About ### About I'm Nasser Ghanemzadeh. 17+ years building products and companies across AI, SaaS, and fintech. Previously CEO/CPO at Nivo (acquired). Currently building Vectig solo with Claude Code, and running a 10-business-day [AI Feature Sprint](https://ghanemzadeh.com/ai-feature-sprint/) for B2B SaaS teams. Also wrote [Forward Deployed AI Engineering](https://ghanemzadeh.com/books/forward-deployed-ai-engineering/) and [Founder Mode](https://ghanemzadeh.com/book/). More on the [home page](https://ghanemzadeh.com/). 07 · Get the next one ### Subscribe One email a week. Built for founders and operators shipping AI-native products in small teams.