SkyPilot
SkyPilot

Forward Deployed Engineer

San Mateo, CA, USOn-siteMidFull-time

Posted 16 days ago

The role in short

Embed with AI teams to deploy and integrate SkyPilot into their cloud/Kubernetes environments, building custom tooling and solving production infrastructure challenges.

About SkyPilot

SkyPilot accelerates the world's most ambitious AI teams. SkyPilot turns fragmented AI compute across clusters into one optimized, highly available and easy-to-use pool: a single AI supercomputer.

SkyPilot (10k+ GitHub stars, 18M+ downloads) manages the GPU fleets of 100s of companies, from Fortune 500s to top AI-natives like Abridge, Applied Compute, Mistral, Unconventional AI, H Company, and Nubank, with usage growing exponentially. Born in the UC Berkeley lab behind Spark and Databricks, our growing team includes top-tier talent from Databricks, Google, Berkeley, MIT, CMU, and Cornell. To date, SkyPilot has raised over $20M in seed funding from top investors (incl. Lux, Coatue, Amplify) and operators (incl. Ali Ghodsi, Jeff Dean, Guillermo Rauch, Amjad Masad, Clem Delangue, Aaron Levie).

The role

SkyPilot powers frontier AI teams and enterprises that run our platform inside their own clouds, Kubernetes clusters, and data centers. We're looking for a Forward Deployed Engineer to embed with them end to end — deploying and integrating SkyPilot, shaping how their workflows map onto it (and how SkyPilot adapts to them), and solving the hard problems that stand between a customer and production. You'll own the technical side of making customers wildly successful. You're the person a frontier team leans on to get SkyPilot running in their own infrastructure.

What you'll do

  • Partner with the teams building frontier AI: work hands-on with foundation labs and AI-natives to architect and deploy their production workloads on SkyPilot — across cloud, VPC, Kubernetes, and on-prem.
  • Lead discovery and sales: design compelling sales demos, run the technical deep-dives that turn a customer's goals into a design they can ship, and build trust with the engineers and technical leaders — CTOs, VPs of Engineering, ML leads — driving the work.
  • Fit SkyPilot to the customer — and the customer to SkyPilot: adapt their workflows to run well on SkyPilot, shape the product to how they work, and turn each engagement into a repeatable playbook.
  • Bring the field back to the product: unblock the hard infrastructure problems in production, and carry customer signal back to the core engineering team.

What we're looking for

  • You've deployed and integrated complex infrastructure or platform software into real customer environments, and you thrive in ambiguity.
  • Strong systems and cloud fundamentals — comfortable across clouds, Kubernetes, and Linux, and able to debug someone else's environment.
  • Strong Python (or similar) — enough to build the glue, tooling, and integrations a deployment needs.
  • Genuinely customer-facing: you communicate clearly, build trust, and turn messy requirements into shipped solutions — you want to be where the product meets the real world.
  • Experience as a forward-deployed or field engineer, familiarity with GPU / ML workloads and the tools to train and serve them, or a developer-facing / open-source background.

What we offer

  • Competitive compensation and equity.
  • Comprehensive medical, dental, vision coverage for you and your dependents.
  • Company 401(k) plan.
  • The chance to work with some of the best minds in cloud, distributed, and AI systems — with significant autonomy and ownership.
  • A front-row seat at the latest open-source infra startup from Berkeley.
  • Gourmet lunch & dinner for the team to do their best work.

Location: San Mateo, CA. Remote will be considered for exceptional candidates.

Prep for this interview

4 questions likely to come up at SkyPilot, each matched to a line in this posting.

  1. 1Cloud, deployment and security
    They wrote: “Partner with the teams building frontier AI: work hands-on with foundation labs and AI-natives to architect and deploy their production workloads on SkyPilot — across cloud, VPC, Kubernetes, and on-prem.”

    A bank wants our product inside their own cloud account, with no data leaving it. How do you approach it?

    How to answer

    Why they ask: Self-hosted and private-cloud deployments are common in regulated industries.

    1. Learn their rules early: network limits, approved services, review process
    2. Package the product so it installs the same way each time (containers, Terraform)
    3. Plan how you'll get logs and updates without touching their data
    4. Run a test install with them before the real one

    Tip: Talk about the conversation with their security team as much as the setup.

  2. 2Ownership and working with customers
    They wrote: “The chance to work with some of the best minds in cloud, distributed, and AI systems — with significant autonomy and ownership.”

    Tell me about something you shipped that broke in production.

    How to answer

    Why they ask: Everyone has broken production. They want to hear you own it and change something afterwards.

    1. What broke and who it affected
    2. How you found out and what you did in the first hour
    3. The root cause, said plainly
    4. What you changed so it can't happen the same way again

    Tip: Pick a real one with a cost. Spend most of the answer on what you changed.

  3. 3Discovery and scoping
    They wrote: “Genuinely customer-facing: you communicate clearly, build trust, and turn messy requirements into shipped solutions — you want to be where the product meets the real world.”

    Tell me about a time the customer asked for one thing but needed another.

    How to answer

    Why they ask: Good FDEs spot when a request is a guess at a solution. They want proof you can redirect a customer without losing their trust.

    1. What they asked for, and the detail that made you doubt it
    2. How you tested the doubt (shadowed users, looked at logs, ran a quick prototype)
    3. How you made the case to the customer
    4. What shipped and what changed for them

    Tip: Pick a story where you checked your hunch with data or users before pushing back.

  4. 4Integrations and APIs
    They wrote: “Strong Python (or similar) — enough to build the glue, tooling, and integrations a deployment needs.”

    Walk me through connecting our product to a customer's internal system that has no documentation.

    How to answer

    Why they ask: This happens weekly in the job. They want a calm, step-by-step method.

    1. Find who owns the system and get read access to a test environment
    2. Watch real traffic or logs to learn the data shapes
    3. Build a thin adapter with logging around every call
    4. Agree on how failures get reported, and who fixes what

    Tip: Talk about finding the owner of the system first. The code is the easy part.

Not the one? Get the next one by email

New roles land here every day. Pick what you want to hear about.

FDE Info

Want the job, not just the posting? Learn how FDEs work.

FDE Info is our free weekly newsletter for forward deployed engineers: tools, field notes from real deployments, and how FDE teams hire and pay.

Free. One email a week. Unsubscribe in one click.

Similar roles at other companies

SkyPilot

On-site

Apply