About the role
Applied AI engineers sit between our models and the problems people actually have. You work directly with customers and internal ventures to scope a use case, build the system around the model, measure whether it works, and hand back something that stays working after you leave.
Frontier AI produces capabilities faster than most organisations can absorb them. The gap is rarely the model alone: it is retrieval, evaluation, tool design, latency budgets and knowing where a human must stay in the loop. What you learn feeds back to research as concrete failure cases, and into products such as Clexerl Go as patterns other teams can reuse.
Engagements are short and hands-on. You write production code, you own the eval set, and you tell customers plainly when a model is not yet good enough for the job.
What you’ll do
- Scope customer problems and decide when a model, a workflow, or plain software is the right answer.
- Build prompt, retrieval and tool-use pipelines with structured outputs, retries and fallbacks that behave predictably in production.
- Design task-specific evaluations, including labelled sets and model-graded checks, and use them to drive each iteration.
- Diagnose failures such as hallucinated tool arguments, context overflow and brittle prompts, and fix them at the right layer.
- Document reference architectures and sample code that other engineers can adopt without you.
- Report recurring gaps to research and platform teams with reproducible examples.
What you’ll bring
- Strong software engineering fundamentals and a record of shipping systems that use language models in production.
- Practical experience with retrieval, embeddings, chunking trade-offs and evaluation of retrieval quality.
- Fluency in Python and at least one other language, plus comfort with APIs, queues and databases.
- Good judgement about measurement: you distrust demos and insist on held-out data.
- Ability to explain technical limits clearly to non-specialists, including when the answer is no.
- Willingness to travel occasionally for on-site working sessions.
Nice to have
- Experience fine-tuning or distilling smaller models for a narrow task.
- Background in a regulated domain such as finance, healthcare or legal technology.
- Experience building agent systems with long-running, multi-step tool use.
- Prior customer-facing technical role, such as solutions engineering or consulting.
Work you might take on
- 01
Build a document-heavy question answering system with citation checking and an evaluation set drawn from real customer queries.
- 02
Design a tool schema for an operations agent and measure how argument errors fall as descriptions and examples improve.
- 03
Create a regression suite that replays logged conversations against new model versions and summarises behavioural differences for the customer.
- 04
Prototype a human-review workflow where the model drafts, flags its own low-confidence outputs and routes only those to a person.
Compensation
$240K – $380K
The US base salary range for this role. Every offer also includes equity in Clexerl and, when it spins out, in the venture you help build. We tell you the reasoning behind your number.
Benefits
Health, fully covered
Medical, dental and vision for you and your dependants, premiums paid in full.
Two kinds of equity
Equity in Clexerl, plus a stake in the venture you help build when it spins out.
22 weeks parental leave
Fully paid, for every parent, however your family arrives.
Rest that sticks
Flexible time off with a four-week minimum we actually check on, plus a company-wide week in December.
Learning and compute
A yearly budget for courses, conferences and books, and compute credits for side projects.
401(k) with a 4% match
Matched from your first paycheque and vested immediately.
How we hire
Apply 5 minutes
Send links to what you have built and a few lines on why this role. No cover letter needed.
Intro call 30 minutes
A conversation with the hiring manager about the work, the team and what you want next.
Work sample 2–4 hours
A paid take-home or a pairing session on a real-shaped problem. You pick the format.
Studio day One day
Meet the team, go deep on your work and ours. Remote or in person.
Offer Within 5 days
A written offer with the full range, the equity and the reasoning behind both.
Opens an email with a short template. Links beat résumés.