off-the-stack
cd ~/careers
Field / hybrid engineeringaka "FDE"

Forward Deployed Engineer

An engineer who deploys into the customer's world and builds the last mile that makes the product real.

Entry
$120k
Mid
$175k
Senior
$260k+
Demand
Rising

Popularized by Palantir and now spreading fast through the AI-startup world, the FDE embeds with a customer, learns their domain, and writes the integration and glue code that turns a generic platform into something that solves the customer's specific problem. It's part consultant, part engineer, and one of the most intense and well-paid hybrid roles in tech.

The myth

It's just consulting with a fancier title.

The reality

You write production code that ships into the customer's environment and feeds the core product. You're an engineer who happens to work at the customer's coalface, not a slide-deck consultant.

cat ./what_you_actually_do.md

  • Embed with a customer, learn their domain fast, and find where the generic product needs a custom last mile.
  • Write real integration and glue code against the customer's messy, real-world data and systems.
  • Build bespoke workflows and tooling on top of the core platform to fit one customer's exact problem.
  • Feed everything you learn at the coalface straight back into the core product roadmap.
  • Own the technical success of a flagship account end to end, often with very little hand-holding.

cat ./why_underrated.md

It's a newer title that hasn't reached most career-advice channels yet, so students and even mid-career engineers simply haven't heard of it. The people who find it tend to love it: you see your code create value almost immediately, you get unusually broad exposure to real-world systems, and you're paid at the top of the engineering range for the intensity. The AI wave made it hotter — every AI startup selling into enterprises suddenly needs engineers who can bridge a powerful-but-generic model and a customer's specific, unglamorous reality.

grep -i 'good fit' ./who.md

  • Engineers who want their work to visibly matter, fast, and don't mind ambiguity.
  • Generalists energized by new domains and real-world messiness.
  • People who like customer contact but want to stay deeply hands-on in code.

cat ./pay.md

FDE comp runs high because the role is hard and the impact is direct — you're often the reason a flagship account succeeds. At AI startups and Palantir-style companies, senior FDEs with equity reach well past $250k all-in. The trade-off is intensity and, sometimes, travel.

./break_in.sh

  1. Show range, not depth

    FDEs are valued for breadth — many languages, many systems, fast learning. A varied portfolio beats one deep specialty here.

  2. Build something end to end with messy data

    Take a real, ugly dataset and ship something useful. That's the job in miniature.

  3. Target AI and data startups

    This is where FDE hiring is exploding. Many will take a strong generalist SWE and grow them into the role.

  4. Lean into customer-facing wins

    Any time you've built something for a real, specific user and watched it land — that story is exactly what they're screening for.

tail -f ./a_day.log

  • 09:00On-site (or on-call) with the customer's team; map a workflow the core product doesn't cover yet.
  • 11:00Write integration code against their real data sources — half the battle is the data being a mess.
  • 14:00Build a custom dashboard/tool on the platform that solves this customer's exact problem.
  • 16:00File precise product feedback so the next customer gets this for free.

ls ./toolbelt

  • Several languages
  • SQL & data wrangling
  • APIs & integrations
  • Cloud platforms
  • The core product/platform
  • Domain curiosity