Forward deployed engineers: in demand, median experience required is just 5 years  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ 
Forward deployed engineers: in demand, median experience required is just 5 years
͏ ‌     ­ ͏ ‌     ­ ͏ ‌     ­ ͏ ‌     ­ ͏ ‌     ­ ͏ ‌     ­ ͏ ‌     ­ ͏ ‌     ­ ͏ ‌     ­ ͏ ‌     ­ ͏ ‌     ­ ͏ ‌     ­ ͏ ‌     ­ ͏ ‌     ­ ͏ ‌     ­ ͏ ‌     ­
The Output by DataCamp

A weekly dispatch for AI + Data Practitioners  ·  #022

Send in the forward deployed engineer. Then pay them $175K–$1M+

 

There are 1,000+ open forward deployed engineer (FDE) roles right now, and the biggest AI companies can't fill them fast enough.

Palantir invented the job with a simple idea: if customers can't make your software work, send an engineer who can. Now OpenAI and Anthropic want their own. The catch? You need to build AI systems and win over the people who'll use them. Do both, and companies will shell out.

What it is. What it pays. What it takes

An FDE embeds with one customer and fixes the problem inside their actual environment: old systems, messy data, and internal politics included.

The entry bar is lower than you'd think. Half of postings ask for five years' experience or less. The pay isn't: $175K–195K base before stock and bonuses, and up to $1.28M at OpenAI.

Hiring managers want proof, not credentials: what you found, how you diagnosed it, and what changed.

Brand new skill track: Forward Deployed Engineer. 10x increase in platform search, $175K–195K average base, 5 years of experience required (median)

Our learners saw it coming

Searches for FDE grew more than 10x between May and October 2026, and mentions of the role to our AI Assistant grew 6.6x.

The number of learners taking AI-building courses paired with software engineering courses within 90 days is up 60% year over year, and 32% in the last six months.

How to build the experience

Start with a real app: an API, a database, and deployment. Add one LLM feature, make it agentic so it calls tools across multiple steps, then add evals, tests that check whether its answers are actually good. Most people skip evals, which is exactly why they'll set you apart.

Our latest track covers the AI core, from your first prompt to production-ready agents, plus how to explain your work to people who've never touched code.

68%

of FDE job postings don't ask for a degree

The customer half, you build yourself. Find one real business with one real inefficiency, like a shop still tracking inventory by hand. Learn how the work gets done, and keep improving your fix until something measurable moves. That's the proof job postings ask for.

Start the Forward Deployed Engineer Track →

New to coding? Start with our Python Programming Fundamentals track: 16 hours of Python for building software and the foundation the FDE track builds on. No experience needed.

Black Friday: 60% Off

Black Friday is LIVE. AI skills pay 62% more. Build them at 60% cheaper, our biggest sale of the year.

Get Premium Now
 

This week in AI and data

Is Huang's exit an NVIDIA problem or an AI problem?
 

AI News

Is Huang's exit an NVIDIA problem or an AI problem?

NVIDIA's CEO challenge, Anthropic tops another ranking with Fable 5.2 while Google and Meta ship cheaper models, three companies announce new AI models for robots, and more.

Read Now
 
Meta's Muse Ember: A low-cost AI model built for agents

AI Agents

Meta's Muse Ember: A low-cost AI model built for agents

Two models, identical pricing. One gets the job done for a third of the cost. The real lesson in Meta's new Muse Ember isn't the model, it's how the work gets split.

Read Now
 
A deeper look: The Forward Deployed Engineer

Career Guide

A deeper look: The Forward Deployed Engineer

Uncover what these engineers do day to day, the specific tasks AI and enterprise software companies need them for, and the skills you need.

Read Now
 

Courses learners love

Claude Fundamentals
Claude Fundamentals

Skill Track

New

Skill Track

New

Claude Fundamentals

7 courses  ·  16 hours

Reliable results take more than a good prompt. Go from your first conversation with Claude to building your own reusable tools, learning to shape Claude around how you actually work.

Start Track
Associate AI Engineer for Developers
Associate AI Engineer for Developers

Career Track

Certification Available

Career Track

Certification Available

Associate AI Engineer for Developers

10 courses  ·  29 hours

If the FDE got you thinking, this is where that engineering half starts. Same core stack: OpenAI API, LangChain, MCP, LLMOps, plus an industry-leading certification.

Start Track
Security in Google Cloud
Security in Google Cloud

Skill Track

Security in Google Cloud

7 courses  ·  26 hours

Built with Google Cloud. Learn how to secure your Google Cloud end-to-end, from IAM and VPC controls to hardened VMs, containers, data, AI models, and defenses against DDoS attacks.

Start Track
The data engineering skills employers are hiring for in 2027
 

The data engineering skills employers are hiring for in 2027

 

In four free live sessions with experts from Snowflake, Databricks, and dbt Labs, build your first AI-assisted pipeline and a working app to show employers.

 
Live Nov 30 – Dec 4
 
Save My Spot

Resources to level up

Code Along

Mon, Dec 7, 11 AM ET

Getting started with Gemini and NotebookLM

A hands-on start with Google's AI tools for research and decision-making, taught through one real-world example: improving your golf game.

Save My Spot
 
Cheat Sheet

Cheat Sheet

Python with AI

Building the portfolio project from the top spotlight? Vague prompts produce fragile code, clear ones don't. The five principles that turn AI-generated Python into code you can actually trust.

Download Now