Location: Palo Alto, CA (5 days/week in office) •
Type: Full-time •
YOE: 2yr+ FDE experiences. (Bonus) technical sales, technical consultant or AI/SaaS founder experience •
Compensation: competitive base + equity
About the roleYou will be one of our first Forward Deployed Engineers, working directly with the founding team. Your job is to shorten the path from "interesting first call" to "signed customer." You will own three things:
- Run technical evaluations that close. Scope, deploy, and run Goaly pilots inside prospect environments, and turn each one into a clear, numbers-backed result the buyer can take to their leadership.
- Be the technical voice in the deal. Join discovery and evaluation calls, answer the hard infrastructure and security questions on the spot, and remove the blockers that stall deals.
- Bring the field back into the product. Turn what you see in real clusters into integrations, fixes, and product changes, so every next deployment is faster than the last.
You need strong systems engineering skills, because you will work inside other people's infrastructure with limited access and real constraints. You need customer instincts, because the job is to make the prospect successful first and Goaly successful as a result. You do not need a background in model research.
This is a founding role. There is no playbook yet; you help write it, and the deployment process you build becomes how Goaly scales to every customer after you.
What you'll doDeploy Goaly in prospect environments- Scope each pilot with the prospect: which clusters, which workloads, what success looks like, and how we will measure it.
- Install and configure Goaly on-prem across the stacks our customers run (Kubernetes, Slurm, Ray, and in-house schedulers), in read-only, telemetry-only mode.
- Work through security reviews, network constraints, and access approvals with the prospect's platform and security teams.
- Debug what breaks in real environments: odd scheduler setups, missing metrics, custom training stacks.
Prove the value and convert- Turn pilot data into a clear business case: compute wasted, compute recovered, and the cost of each model improvement.
- Present results to ML leads, infra leads, and the executives who sign.
- Partner with GTM on deal strategy, technical objections, and the path from pilot to paid deployment.
- Keep each evaluation on a tight timeline, and flag early when a deal is at risk.
Build what the field needs- Write the integrations, scripts, and connectors that a deployment needs, and push the reusable ones back into the product.
- Build and maintain demo environments and reference deployments for sales calls.
- Turn repeated questions into docs, runbooks, and install guides, so time-to-value drops with every customer.
Feed the roadmap- Bring precise, prioritized field feedback to engineering and the founders: what customers need, what blocks adoption, and what wins deals.
- Reproduce customer issues and work with engineering until they ship a fix.
What a typical week looks like- Two days on active pilots: on-site or remote in a prospect's cluster, installing, debugging, and pulling results.
- A day on deal calls: discovery, technical deep dives, security reviews, and results readouts.
- A day building: an integration a prospect needs, a better install path, or a demo environment.
- A day with the team: engineering sync on field feedback, GTM sync on pipeline, and founder time on what we learned.
- [Roughly X percent travel] for on-site deployments and customer visits.
You may be a good fit if you have- A strong engineering background: [2-5] years as a software, infrastructure, or ML engineer. You write production-quality code (Python required; Go or Rust a plus).
- Hands-on infra experience: you have run or debugged workloads on Kubernetes, Slurm, or Ray, and you are comfortable on Linux, in cloud and on-prem environments.
- Customer-facing technical experience: as an FDE, solutions engineer, sales engineer, or similar. You have taken a customer from a blank page to a working result.
- Clear communication: you can explain a technical result to an engineer in the morning and to a VP in the afternoon.
- Ownership under ambiguity: you drive a deployment to the finish line without a detailed spec or someone checking in.
Strong pluses- Experience with GPU clusters and distributed training (PyTorch, DeepSpeed, FSDP, Megatron, or similar).
- Experience with observability and telemetry (Prometheus, Grafana, DCGM, OpenTelemetry).
- Experience at an AI infra, MLOps, or developer-tools company with an enterprise sales motion.
- Experience with enterprise security reviews and on-prem software deployments.
- Startup or founder experience.
Reasons to take this job- You work on the most compute-intensive workloads in the industry, inside the clusters of the teams that run them.
- You see the direct impact of your work: every pilot you run is a deal that closes or doesn't.
- Your feedback goes straight into the roadmap, with no layers between you and the founders who built post-training infra at Llama scale.
- You build the FDE function from 0 to 1, with a clear path to leading it as it grows.
- Great vibe: a strong (and fun) team who worked at the frontier before, substantial GPU capacity to test ideas, and real customers to build with.
Reasons to not take this job- You want to work only on the core product, not inside customer environments.
- You don't want to be on calls with customers, or to travel for on-site deployments.
- You want a narrow scope. This role moves between engineering, deals, and product every week.
- You want a playbook. There isn't one yet.
Location, visa sponsorship & benefits- Hybrid in Palo Alto: 4+ days/week in office.
- Visa sponsorship: H-1B and OPT/CPT support available, with immigration counsel.
- Meals & perks: Complimentary lunch, dinner, snacks, and drinks.
How to applySend us 2 links or short write-ups to start: something you built or deployed in a customer environment, and a technical problem you solved under real constraints (limited access, a deadline, an unhappy customer). If you can't share details, tell us in a paragraph how you would run a 2-week Goaly pilot on a prospect's 500-GPU cluster.
A note on qualifications. We value exceptional ability over perfect keyword matches. If the work excites you and you can show strong technical ability, customer instinct, and ownership, we encourage you to apply.