We help companies rethink how work gets done when intelligence becomes cheap, abundant, and programmable.
We value engineers who recognize that AI is a tool rather than a universal remedy. The most effective builders reshape fuzzy work into clear systems: they define inputs, specify success criteria, expose hidden judgment, build tool loops, author evals, and designate where a human must approve the next action. We hire for raw problem-solving speed and steep growth curves.
We do not require a minimum number of years. Show us that you reason quickly, build competently, and think from first principles. If your growth trajectory is steep, your upside is substantial.
The role is in-person in Bangkok. We work 5-6 days a week.
What AI-enabled means here
You prompt, steer, critique, and iterate with AI tools effectively. You distinguish an impressive demo from a reliable solution.
You understand model capabilities and boundaries: what models solve well, where they hallucinate, when they need tools, and when human judgment must own the decision.
You evaluate open-source and frontier models across capability, cost, latency, privacy, tool support, and operational risk.
You think in evals. Before you trust automation, you ask what proves it works, what proves it fails, and which test cases belong in the regression suite.
Open roles
Pick a role below, then email [email protected]. Share concrete proof of your work: a project, demo, experiment, workflow, eval harness, technical write-up, or effective AI-built tool.
Role 1
AI-enabled engineer
Turn messy operational work into reliable code, command surfaces, structured context, and governed AI workflows.
You will build leverage: tight systems that absorb recurring tasks, make judgment explicit, and guide operators reliably from raw inputs to reviewed outputs.
You will work on
Build internal tools that convert recurring operational tasks into repeatable CLI commands, scripts, schemas, and evidence packets.
Develop AI product surfaces where models draft, retrieve, transform, and propose, while human operators approve risky or consequential actions.
Author comprehensive eval suites for AI workflows: test fixtures, golden baselines, adversarial edge cases, traces, and reviewer dashboards.
Integrate lightweight pipelines across APIs, inboxes, calendars, chat platforms, notebooks, and internal company databases.
You are probably strong if
You prototype fluently in TypeScript or Python and read unfamiliar codebases quickly with AI as your pair.
You know when to prompt a model, when to attach tools, when to inject retrieval context, and when to simplify the problem with deterministic code.
You evaluate the trade-offs between hosted frontier models and open-source models with engineering rigor rather than ideology.
You make systems legible: you document what a tool does, where it fails, and how metrics prove it improved.
Apply
Email [email protected] with the subject "AI-enabled engineer". Include one concrete system you built or evaluated. We value demonstrated ability over polished credentials.