Founding Full-Stack Engineer

Description

The company is an AI-native law firm. We don't sell software to lawyers; we are the law firm, and our engineers build the agents that do much of the work. Every improvement to our agents shows up directly in our margins and our clients' bills.

We're not trying to make lawyers 20% faster. We're building a law firm where software performs most of the work and our lawyers own the result. Our agents do the drafting and review; our lawyers spend their time on the judgment. That changes the engineering problem. Our agents have to be reliable enough for a lawyer to put their name on the output, and for a client to act on it.

The opportunity

Most companies building AI for professional work stop at copilots. We're going much further: systems that perform large parts of the work of a professional services firm. Our lawyers own the outcome, That means solving problems that are still wide open. How do agents operate across long-running workflows? How do you evaluate work where mistakes matter? When should a system act, and when should it escalate? How do humans and agents divide thousands of consequential decisions every day?

If we're right, AI-native professional services firms will look more like software companies with professionals in the loop.

About the role

You'll join our founding engineering team in San Francisco, working in person alongside the founders and our lawyers. Our lawyers sit next to you and use what you build all day. When an agent gets a redline wrong, you hear about it that afternoon and fix it together.

We've built agents that draft, review and negotiate contracts, intake that turns client requests into structured legal work, and software used every day by lawyers and clients. Your mission is to iterate and improve on that system, to handle hundreds, and eventually thousands, of matters a day without compromising the quality.

Who you'll work with

Pamir, our CEO, advised OpenAI on contracts and privacy. Daniel, our COO is a Harvard-trained lawyer who has worked with frontier model labs. Stefan, our CTO, did his graduate work at MIT and has scaled ML products to hundreds of millions of users. Our lawyers come from Fenwick, Cooley and Goodwin.

What you'll own

Shaping features end to end. Sit with our lawyers, understand the problem, build it, ship it and get feedback the same day. Building functional and beautiful client surfaces for intake, communication with counsel and delivery of results. Work together with design engineers to ship intuitive, sleek and responsive pages. Iterating on the lawyers' agentic workspace: review queues, redline and diff views, agent chat surfaces, citations, and controls to approve, edit or escalate in seconds Building operations tooling including dashboards, automations and integrations with Slack, email, document and e-signature platforms, and payment and billing providers Making our agents trustworthy and reliable at scale. Improve the the evals, observability and infrastructure that ensure agents are doing the right thing Take part in shaping how we build. Architecture, engineering standards and the team around you.

How we build

We're TypeScript end to end: a Next.js and React frontend on a tRPC API, Node.js and PostgreSQL on AWS. Our agents run inside harnesses we build ourselves, orchestrated as durable workflows in Temporal, powered by popular agent SDKs.

Responsibilities

• Are a product-minded full-stack engineer who likes owning problems from the database to the UI.
• Shipped agentic systems to production and can share some war stories
• Maintained and evolved a large codebase over years at a company past the prototype stage
• Want the intensity of an early-stage company: priorities move quickly, you work directly with users, and you carry more responsibility than you would anywhere larger.
• What you get
• Competitive salary and significant equity, in a company still early enough for your work to materially shape the outcome.
• The best AI tooling we can buy, and a code base optimized for agentic development.
• A real say in defining what an AI-native professional services firm looks like.
• If making AI reliable enough to do work people currently pay lawyers hundreds of dollars an hour for sounds like an unusually interesting engineering problem, we'd like to talk.
• Skills: Amazon Web Services (AWS), PostgreSQL, React, TypeScript, Next.js, AI Agents
• Experience: 3+ years
• Visa: US citizen/visa only