👀 Google Cloud announced it's hiring hundreds of Forward Deployed Engineers - and this isn't a recruiting press release. Behind it is a U-turn in the business model: from "we sell access to Gemini" to "we send our engineers to live inside the client's processes until the AI actually works".
What is a Forward Deployed Engineer:
Google opened 59 roles across the US, London, Paris and Hong Kong. An FDE isn't a consultant who gives a presentation and leaves. It's a developer who gets dispatched inside the client: writes production code right in their infrastructure, iterates until the system runs stably in prod, brings the patterns they find back to Google's product team. The difference from classic consulting is fundamental: an FDE is accountable for what works in production - not for what's written in the final report. Base $127K-$183K + equity, senior packages reach $400K+
The model was invented by Palantir back in 2003 - it made money by literally sewing its engineers into the processes of the CIA and big corporations. Now Google, OpenAI and Anthropic are copying the formula, only instead of intelligence data - Gemini, GPT-4o and Claude.
Why FDEs are even needed - what's the problem:
95% of corporate AI pilots fail. Not because the models are bad - because the gap between what an AI lab can do and what a bank's or retailer's IT department can do is huge. The client understands its business, Google understands how Gemini works in prod. Neither side has the other's expertise. An FDE closes this gap physically - by being right there on site
What Google is reshaping through FDEs:
💡 Kills "sell and forget" - the engineer stays until the business metrics actually move
💡 Plugs in through integrators - $750M into the partner ecosystem, FDEs sit inside Accenture, Capgemini, Deloitte, PwC, TCS
💡 Feeds patterns into the product - everything an FDE finds inside the client goes back into Gemini
💡 Has token leverage - Google Cloud API: 16 billion tokens per minute, up from 10 billion in a quarter
The FDE race is already on:
⚡️ OpenAI - bought Tomoro (~150 specialists) and set up an AI consultancy with Bain, Capgemini and McKinsey
⚡️ Anthropic - building up embedded teams for enterprise
⚡️ Palantir - the original of this model, 20 years on the market, watching with a smile
⚡️ Microsoft and AWS - their own versions for Azure OpenAI and Bedrock
⭐️ Thomas Kurian, CEO Google Cloud:
"If you are a builder who wants to work on the world's largest stages and be at the center of the agentic era - join us"
🎯 Palantir invented this model and became a $100B+ company. Now all the AI giants are in the same game, only with LLMs inside. We're leaving the SaaS era and entering the era of "AI as an embedded team". And here's the interesting part for developers: FDE is the next career level in AI. Not the person who writes prompts, but the one who can take a model's black box and make it work in a corporation's real processes. Judging by salaries up to $400K - the market already gets it 😎