About this Data Engineer (3-6 month contract) role at SeedLegals
SeedLegals is the leading provider of automated legal solutions for startups, with offices in London, New York and Paris. Since launching in 2016, we’ve supported over 75,000 startups who have collectively raised more than £2bn on our platform.
We’re a Series A company backed by Index Ventures, on a mission to make entrepreneurship more accessible and scalable.
Our values are key to our success; here at SeedLegals, we:
- Invest and trust in each other;
- Are committed to a growth mindset;
- Embrace diversity and cultivate inclusion;
- And are driven by customer success.
About the Role
We're looking for a contract Data Engineer to join us for 3–6 months to build a new AI-powered BI assistant from the ground up. This is a dedicated, standalone project — you'll own it solo, from architecture decision through to a working, deployed agent.
Much of our data team's time today goes into one-off reporting requests — dashboard and number-pulling work that doesn't always need a human in the loop. Your job is to build the agent that takes that work on: an AI assistant, most likely Claude-based and connected via MCP to our warehouse (BigQuery/dbt), Metabase, and GitHub, that answers routine BI questions end to end and flags the complex ones for human review. You'll score it against an evaluation set of real historical business questions — on accuracy, run cost, and maintainability — and use that to guide the build, before deploying to an internal pilot group.
This is a rare contract: a genuinely greenfield agentic AI build with a crisp, measurable deliverable, inside a company at the heart of the UK startup ecosystem.
About the Team
We're an international team of around 20+ developers from 12 countries based in the UK and France, with a hybrid working model that allows for flexibility and collaboration. While each team meets fortnightly at our offices near Old Street to whiteboard, plan, and socialise, you'll also have the freedom to work remotely.
Requirements
We'd love to hear from you if...
- You love solving challenging problems
- You have strong SQL and dbt experience, and you're comfortable working within an existing real-world lineage rather than greenfield modelling
- You have hands-on experience building LLM/agent systems (Claude, MCP, or similar) — this is the core of the role, not a bonus
- You love working with AI agentically
- You're comfortable in Python for tooling and integration work
- You're happy owning a project solo, end to end, without a team to lean on day-to-day
- You own and take pride in contributing to all stages of development — from architecture through evaluation to deployment
And if you can hit one or two of these as well, even better...
- You're familiar with Metabase or similar BI tools
- You've worked with YAML-based BI/modelling tools (e.g. Bruin)
- You've written your own plugins, skills or agents for LLMs such as Claude
- You've built evaluation sets for LLM systems and thought hard about accuracy vs run cost
- You have some personal projects on the go (a blog, open source contributions, side builds etc.)
Benefits
You'll join a friendly, motivated and multicultural team of founders, investors, developers and growth hackers, based near Old Street in the heart of London.
- A well-scoped, standalone project with a clear deliverable — no BAU distractions
- Hybrid/remote flexibility with London office access
- Free lunch in the office once per week
- Extensive exposure to the startup world: VC fundraising, investor tax schemes, and the tooling behind 75,000+ startups
Our Interview Process:
- 30min initial call (video call)
- 1hr technical interview — we'll walk through the project brief and dig into your experience building agentic/LLM systems and working with production data models
- 30min wrap-up call with Anthony Rose, our CEO
