Founding Full-Stack Engineer, AI Agents

TLDR: This isn't a regular engineering job given the autonomy, pace and progression path available. We’re looking for a full-stack engineer to help build an agentic commercial insurance brokerage from the ground up. Venture-backed team comprising repeat founders and experts in insurance technology and broking.

Why Vara

Businesses buy over $1 trillion of commercial insurance every year, and around 90% of it is still run by hand. At Vara, we’re building a UK commercial insurance brokerage where we’d like 90% of it to be run by agents.

We’re doing that in two ways. First, we’re making the professionals who help businesses buy insurance 10x more efficient than they are today. Second, we’re building for a world where businesses start their insurance journey with AI: Claude, ChatGPT or Grok, or a personal assistant like Muse or Instinct that already knows their business.

You’ll have extreme ownership and autonomy building the AI operating and distribution system that will reshape a trillion-dollar industry, alongside a low-ego, high-velocity team.

What you’ll do (key list below but be ready to adapt to what is required given you will be the tech stack):

• Own products end to end: work with our brokers, shape the requirements, then build the frontend, backend, agents and evals.
• Build for AI-first distribution: connectors, structured APIs and MCP integrations so that AI platforms and personal assistants can ask the right questions and get a quote from Vara.
• Turn broking expertise into code: risk gathering, submission packaging, market comparison, insurer chasing, renewal prep.
• Build voice agents that take client and insurer calls and hand over cleanly to a broker.
• Own the number, not the feature. "Quote turnaround went from three days to four hours" beats "we shipped the thing."

Who you are:

• 5-10 years' experience
• A full-stack engineer who has built production systems that real people depend on
• You’ve shipped agentic systems to production, not demos (side projects count)
• You have sharp product instincts and would rather watch a broker work than guess
• You’re fast and passionate about using AI coding tools
• You care more about commercial impact than elegant code
• You’re willing to learn insurance properly

Experience level matters less than the quality of your thinking and what you’ve shipped.

Bonus: MCP and agent integrations, voice AI, RAG, evals, document extraction, early-stage experience.

The honest bit: insurance is unglamorous and full of detail that matters, with insurer portals that are decades old. We’re regulated, so some things have to be right, not roughly right. And we’re early: small codebase, few conventions, not many people to ask at 6pm.

Base salary plus equity will be discussed in first call.