AI Product Engineer
First seen by Alion on Oct 13, 2026.
SmallClaims is a legal tech company that recovers small, valid claims of up to €25,000 through the courts. Claims that businesses write off today because pursuing them costs more than they are worth. Our AI agents run the whole process, from legal assessment and settlement to filing, judgment and enforcement, while our in-house lawyers quality-assure the work and appear in court. That has cut manual case handling from over 15 hours to about 1.5 hours per case. We have handled 5,000+ cases, win 90%+ in court and serve 500+ Danish customers. Case intake has grown more than 370% since 2023, and we have just raised €2M from Vendep Capital to grow in Europe. Now we are growing the engineering team.
You take ownership of your own areas within our AI service (Python/FastAPI, LLM pipelines) and our back-office platform, the tool our case managers and lawyers run the operation on. Ideas take form together with the team, product specialists and the CTO. You take them from there, end to end across the stack: shape, build, ship, verify, follow up.
Own parts of and build features in our AI service: LLM orchestration with structured outputs, tool calling and human-in-the-loop design.
Build the back-office platform our case managers and lawyers use every day, and talk to them to find out what they really need.
Set up and improve evals and observability, so we know our agents are right before we scale them.
Ship and run what you build, from staging to production on AWS, with tests and CI as a habit.
Stack: Laravel/PHP, Next.js, Python/FastAPI, PostgreSQL, AWS, Google.
How we build: Coding agents are part of daily work here. We expect you to be ahead of the curve on building with AI, not catching up to it.
3+ years building and operating software in production, from a startup, a bigger company, open source or other serious project.
You think like a product owner. You want responsibility and ownership, and you take a complicated problem to a shipped feature.
You hold strong opinions lightly. You are direct in disagreement, but easy to work with after the decision.
You communicate directly and are honest about mistakes. You can talk to real users and are comfortable with shifting priorities.
You love building, love learning, and are curious about the domain itself: the legal process, not just the tech.
Must have
Strong backend development, incl. background jobs, auth models and API design.
Solid relational database experience (Postgres/SQL or similar): schema design, migrations, performance.
Hands-on LLM building in production, not just demos: orchestration, structured outputs, tool calling, evals, tracing, prompt design, injection.
Works daily with AI coding agents (Claude Code, Codex, opencode or similar), knows their strengths and limits, and stays up to date.
Testing as a habit: feature and end-to-end tests, CI, verifying your own work through staging to production.
Cloud and deployment literacy (AWS or similar, Docker, CI/CD). You can operate what you ship.
Nice to have
Python/FastAPI, Laravel/PHP or Next.js specifically.
Production migration: third-party service changes, database moves, major-version upgrades, data backfills.
Operating AWS (EC2, RDS, S3, IAM, CloudWatch) and Linux servers hands-on.
Eval or observability tooling for agentic workflows.
Integrations where no documented API exists: webhooks, browser extensions, automation, scraping.
Legaltech, fintech or another regulated domain.
Real ownership of core parts of the AI platform in a small team that ships every week.
Close work with the CTO, product specialists and the lawyers who use what you build.
Our base is the office in Copenhagen, but regular remote work is fine
Apply via The Hub with a short intro, your CV, GitHub and links to things you have built: apps, side projects, something you shipped at work. Next, we send you our skills checklist as a self-assessment plus a few short written questions, then interviews with the team, then an offer. We keep the process short and give you a clear answer at every step.