Make AI do real work.
Most companies have an AI strategy. Very few have AI running in production. We build the systems that close that gap, inside the businesses that use them, with results they can measure.
The gap is the job.
Every leadership team we meet has the same story. They bought licences, ran a pilot, saw a promising demo, and then nothing changed. The service desk is still buried. Finance still closes the books by hand.
The models are good enough. What's missing is the work in between: understanding the workflow, connecting the systems, handling the edge cases, earning the trust of the people who have to use it, and proving it paid off. That work is what Lakely does.
Who does well here.
You'll probably like it here if
- You want to see your code in production at a real company within weeks, not quarters
- You like talking to the people who use what you build
- You can hold a high bar for quality and still ship on time
- You'd rather own a problem end to end than hand off a piece of it
It's probably not for you if
- You want a long spec before you start
- You'd prefer to stay far from customers
- You need a large team and established process to do your best work
Open roles
All remote, all paid in USD, all with equity. Don't see your role? Tell us what you'd build here. We read every note.
Field AI Engineer
You're the engineer a client meets on day one and the one still there when the use case is live. You take a goal like "our service desk is drowning" and turn it into a working system inside their stack, with numbers that prove it worked. Experience shipping in regulated industries is a strong plus.
What you'll do
- Map the client's workflow, data, systems and constraints before anyone writes code
- Turn each use case into a plan with clear success metrics and honest trade-offs
- Build hands-on, from first prototype to production, inside the client's environment
- Agree on acceptance tests with the client so "done" means the same thing to everyone
- Train the people who use it and stay close until it sticks
What you bring
- 5+ years in software or technical delivery, with real time spent face to face with customers
- At least one LLM-powered system you shipped to real users
- Production code in Python or TypeScript across backend and front end
- The confidence to push back on a client and keep their trust
- Writing that a CIO and a staff engineer can both follow
Applied AI Engineer
You build the AI inside our use cases: retrieval that finds the right policy, extraction that reads a thousand invoices correctly, an agent that triages a queue on its own. You care more about month six than demo day.
What you'll do
- Ship AI features end to end, from model calls and pipelines to the screen people use
- Build retrieval, extraction, classification and agent workflows tuned to each client's data
- Pick the right model for each job, including when a smaller or fine-tuned one wins
- Write evals that catch regressions before clients do
What you bring
- Strong full-stack skills across web frameworks, backend services and cloud
- Shipped features to real users in production
- Backend systems that hold up under load and messy inputs
- Daily, opinionated use of AI coding tools
Data & Context Engineer
An agent is only as good as what it can see. You build the data layer that gives it the right context at the right moment: clean, current, permission-aware and traceable.
What you'll do
- Bring data from CRMs, ticketing tools, file stores and warehouses into one governed layer
- Match, deduplicate and reconcile records from conflicting sources
- Design data for retrieval: chunking, indexing, metadata and freshness rules
- Respect the client's access rules so an agent never sees what its user couldn't
What you bring
- Experience building and running production pipelines as a data or analytics engineer
- Strong SQL and Python, plus Airflow, Dagster, dbt or similar on a cloud warehouse
- Experience making data from many sources usable
- Regular use of AI tools in your engineering work
How we hire
1Apply
Send a resume or a LinkedIn profile. A link to something you built counts for more than a cover letter.
2Intro call
A conversation with the hiring lead about what you've done, what you want next and whether this role fits.
3Technical deep dive
Walk us through a system you built. We'll ask why you made the choices you did and what broke.
4Work session
A realistic problem drawn from our client work. Use AI tools the way you would on the job, then show us your reasoning.
5Team and founder conversations
Meet the people you'd work with every day, ask anything, and talk through how we run the company.
6Offer
A clear offer with salary, equity and start date, and time to decide without pressure.
Companies have run their pilots.
Now they want AI that runs the work. If you want to build it, we'd like to meet you.