Forward Deployed Engineer (FDE): The hottest AI role of 2026 and how to break in without a CS degree

A new role just appeared, and the demand curve is vertical

A year ago almost nobody had heard of a Forward Deployed Engineer. Today it is the role Salesforce, OpenAI, Anthropic, Google, Palantir, and Databricks are competing for. Indeed went from 643 open FDE roles in April 2025 to 5,330 in April 2026, a 729% jump in a single year. Live Data Technologies tracked it even higher at 1,165% year-on-year.

The supply side did not move that fast. Candidate pools grew about 50%. So this is a hot market where the people who position correctly right now get paid like senior engineers without competing in the senior engineering pool.

This guide answers four things:

1. What an FDE actually does day to day
2. Why every AI company is suddenly hiring them
3. Who is getting hired (including people without a CS degree)
4. A 30-day plan to position yourself for one of these roles

If you want the short version, comment FDE under any of my videos and I will DM you this guide directly.

What is a Forward Deployed Engineer

An FDE is a software engineer who works embedded inside a customer company. Not in a call centre. Not on a Zoom from headquarters. Sitting with the customer's team, learning how their business actually runs, and shipping AI workflows that solve their specific problems.

Think of it as three jobs stacked into one:

• Software engineer. You write production code that runs in the customer's environment.
• Consultant. You sit in meetings with VPs and engineers, figure out what they actually need (rarely what they ask for), and translate it into a technical plan.
• Product manager. You decide what to build, what to cut, and what to ship next.

Palantir invented the role over a decade ago. They sent engineers into governments, banks, and hospitals to deploy software directly inside the customer's stack. It worked so well that Palantir's growth was largely powered by these embedded engineers. Now every AI company is copying the playbook, because shipping a model is easy and shipping a model into a 50,000-person bank is hard.

The shorthand: FDEs are the people who bridge "the AI works in a demo" and "the AI is running in production at a Fortune 500."

The numbers behind the boom

Three data points tell you why this role is suddenly everywhere.

1. Salesforce committed to 1,000 FDE hires. Largest single corporate commitment in the market. The team is anchored on Agentforce and Data Cloud. Salesforce launched the team in April 2025 and built a six-week onboarding program called Ready in Six to scale headcount fast.

2. Job postings grew over 800% in nine months. Between January and September 2025, FDE postings climbed by more than 800%. Salaries cleared $300,000 at the senior end, with OpenAI and Google reportedly going higher to lock in talent.

3. Deloitte says the AI skills gap is the number one barrier to AI adoption. Deloitte's State of AI in the Enterprise 2026 report surveyed 3,235 enterprise leaders. Skills shortage came in as the single biggest blocker. Only 20% of organisations describe their talent as highly AI-prepared. 82% of leaders run some AI training, and 59% still report a skills gap.

That last point is the whole story. The models work. The tools are ready. The bottleneck is humans who can walk into a company, see where AI fits, and build the thing.

The companies hiring FDEs right now

This is not a niche role at one company. The major hirers:

Company	Notes
Salesforce	1,000 FDE commitment. Multiple openings on Agentforce. Apply through careers.salesforce.com (search "Forward Deployed").
Anthropic	Applied AI team. Embeds with strategic customers. Active req on Greenhouse.
OpenAI	FDE team launched early 2024, expanded to ~50 engineers in 2025.
Palantir	The original. Still the gold standard FDE training ground.
Google Cloud	Hiring aggressively for AI agent deployment.
Databricks, Cohere, Ramp, Rippling, Intercom	All running FDE programs.
YC AI startups	Smaller teams hiring 1-3 FDEs to push their product into Fortune 500 customers.

Most of these roles are US-based but a meaningful slice is remote-eligible. Anthropic, OpenAI, and Salesforce all hire FDEs in EMEA, India, and Asia. Salesforce specifically has FDE postings in Hyderabad, Bangalore, and Tokyo. South African and African applicants get the strongest shot through the EMEA-remote roles and through startups that hire globally by default.

What an FDE actually does in a week

The skill mix is unusual because the day is unusual. A typical week looks something like this:

flowchart LR
  A[Monday: meet customer team] --> B[Tuesday: extract real requirements]
  B --> C[Wednesday: prototype the AI workflow]
  C --> D[Thursday: present to VPs]
  D --> E[Friday: ship to production]
  E --> F[Repeat with next customer problem]

The hard part is not the code. The hard part is figuring out what to build. Customers say "we need an AI chatbot" when what they actually need is a workflow that summarises 2,000 support tickets a week into a triage queue. An FDE's job is to hear the first sentence and deliver the second one.

The Silicon Valley Product Group definition captures it well: an FDE is what happens when you take a strong software engineer and give them the context that usually only product managers and consultants have.

Who is actually getting hired (and this is the part that surprises people)

The script everyone repeats is "you need to be a senior engineer at FAANG to break in." That is half right. Senior engineers do get hired. But the role is so new that companies are also hiring from non-traditional paths because the candidate pool is too small.

People landing FDE roles in 2026 come from five backgrounds:

1. Traditional software engineers with 3-5 years of experience. Default path.
2. Consultants from McKinsey, BCG, Bain, Deloitte, Accenture who can already do customer discovery and have shipped at least one technical side project.
3. Product managers who have written production code at some point and are tired of writing tickets for someone else to build.
4. Operations and customer success leaders at SaaS companies who deeply understand customer workflows and have learned to code with AI tools.
5. Self-taught builders who have shipped something real to real users. Could be a SaaS side project, a Chrome extension, a Discord bot, an internal tool used by hundreds of people.

The pattern across all five: they can hold a conversation with a CFO and write the code in the same afternoon. That hybrid skill is rarer than either skill alone, which is why the salary is high.

If you are coming from a non-engineering background, the bar is not "be the best coder." The bar is "build proof you can ship a thing that solves a real problem for real users."

The 5 skills every FDE needs

Salesforce published the skill rubric they hire against. It is the cleanest framework I have seen.

1. Consulting-grade discovery. Walk into a meeting, extract the real requirement (not the stated one), write a crisp summary afterward, push back on scope without losing the room.
2. Full-stack technical chops. You do not need to invent the next transformer. You need to ship production code that calls an LLM API, processes data, hits a database, and renders a UI. Modern AI tooling (Claude Code, Cursor, v0) makes this much faster than it used to be.
3. AI product instincts. Pick the right model, write the prompt, build the eval, know when to use RAG vs fine-tuning vs agent loops. This is where most engineers are weak and most product people are weak in a different way.
4. Stakeholder communication. Present technical progress to non-technical executives. Manage expectations. Build trust. If you cannot read a room full of VPs, the role eats you alive.
5. Comfort with ambiguity. Every customer is different. There is no playbook. You will be the most senior person in the room on AI even when there is a CTO across the table.

Notice that 1, 4, and 5 are not technical. The reason consulting and product people break in is that they already have those three.

Two paths in

Path A: You can already code

You have 2+ years of software engineering experience. The gap to close is the AI side.

Step	What to do
1	Ship a project that uses an LLM in production. Not a demo, not a tutorial follow-along. Something with real users (even 10 is enough).
2	Write up the project on a public blog or LinkedIn. The write-up matters more than the code. Hiring managers skim for storytelling, not for clean commits.
3	Learn evals. Read the Anthropic prompt engineering docs and the OpenAI evals cookbook. Most engineers skip this. It is the #1 thing FDEs are tested on.
4	Apply to AI startups first, then big companies. Startups will reply faster and the work is closer to the FDE day-to-day even if the title is different.

Path B: You cannot code yet (or coded years ago)

You are a consultant, PM, ops lead, marketer, or student. The gap to close is the technical side.

Step	What to do
1	Pick one AI tool stack and go deep. Suggested: Claude or GPT API + Cursor + Vercel + Supabase. That stack covers 80% of FDE work.
2	Build three small projects that solve a real problem. Examples: a Slack bot that summarises your team's standups, a tool that reads PDFs and answers questions about them, a script that scrapes your competitor's pricing weekly. Each should take 1-2 weeks.
3	Document the build process publicly. LinkedIn, X, a blog. Show the customer problem, the build, the result. This is your proof.
4	Apply to Applied AI / Solutions Engineer / Forward Deployed Engineer roles. Frame your background as "domain expertise + AI execution," not "career switcher."

Tip: If you are in South Africa or anywhere in Africa, prioritise remote-first companies and US/EU startups that say "global hire" in the job description. Salary arbitrage works strongly in your favour at FDE pay bands.

Where to apply this week

Direct links to the actual job boards:

• Anthropic Applied AI — job-boards.greenhouse.io/anthropic (filter for "Forward Deployed" or "Applied AI")
• OpenAI — openai.com/careers
• Salesforce — careers.salesforce.com (search "Forward Deployed Engineer")
• YC Work at a Startup — workatastartup.com (filter for AI, then look for FDE or "applied AI engineer")
• Wellfound (formerly AngelList) — wellfound.com (filter for remote AI roles)

Two job titles to also search, because the role goes by different names: "Applied AI Engineer," "AI Solutions Engineer," "Customer Engineer (AI)." Same role, different labels.

Salary range you should expect

Numbers from Live Data Technologies and BigGo Finance reporting:

• Entry / mid-level FDE: $150k-$220k base in the US, R600k-R1.2m if remote-from-SA
• Senior FDE: $220k-$300k base, plus equity
• Principal / Lead FDE: $300k+ base, with packages reportedly clearing $500k all-in at top labs

For African applicants on EMEA-remote contracts, the upper bound is lower (companies adjust for cost of living) but still represents 3-5x the local senior engineer salary at most SA tech companies. The maths is good even after the adjustment.

The 30-day starter plan

Stop reading career advice. Start building proof.

Week 1: Pick a real problem. Look at your current job or your university and find one workflow that wastes hours. Examples: weekly reports that get manually compiled, customer support tickets that get manually routed, lecture notes that nobody re-reads. Write the problem in one paragraph.

Week 2: Build the rough version. Use Claude or ChatGPT + Cursor. Aim for ugly and working, not pretty and broken. The deliverable is a tool that you (or one other human) actually uses.

Week 3: Get 5-10 real users. Show it to colleagues, classmates, friends in similar roles. Gather feedback. Make 3 changes based on what they tell you. Real users transform a hobby project into a portfolio piece.

Week 4: Write it up. LinkedIn post or short blog. Structure: the problem, who has it, what you built, what you learned, what you would do differently. This is your application material. Now start applying.

Done when: You have a public link showing one shipped tool, one written narrative about how you built it, and applications out to at least 10 FDE-adjacent roles.

The South African angle

Three things working in your favour if you are applying from SA or anywhere in Africa:

1. Remote AI hiring is normalised post-2022. Salesforce, Anthropic, and OpenAI all have global hires.
2. The salary arbitrage is real. Senior FDE roles pay 5-10x local equivalents.
3. Time zone overlap with EU is clean. SAST gives you full-day collaboration with London, Berlin, Paris.

Two things working against you:

1. Visa sponsorship for US-based FDE roles is competitive. Target remote-first companies first, sponsorship-companies second.
2. The senior engineer-on-paper expectation. Compensate with strong proof-of-shipping. Public projects beat unverifiable resume claims.

Sources I used to fact-check this

• Salesforce, "Forward Deployed Engineers Are Proving AI Makes Tech Jobs More Human" — salesforce.com/news/stories/forward-deployed-engineers-making-ai-jobs-more-human/
• Deloitte, "The State of AI in the Enterprise 2026" — deloitte.com/us/en/what-we-do/capabilities/applied-artificial-intelligence/content/state-of-ai-in-the-enterprise.html
• BigGo Finance, "AI Commercialization Fuels 800% Surge in Forward-Deployed Engineer Jobs"
• Paraform, "Forward-Deployed Engineers: How Demand Grew 10x in 18 Months"
• Pragmatic Engineer newsletter, "What are Forward Deployed Engineers, and why are they so in demand?"

What's next

• Comment FDE under any of my videos and I will DM you this guide
• If you want help picking which AI stack to learn first, comment STACK for my exact AI builder stack
• If you want to know which skills are getting hired in SA right now (not just FDE), comment DEVSA

This guide powers the "Comment FDE" CTA across all platforms.