Founding Legal Engineer

Location: Santa Monica, CA (hybrid / in-office preferred)
Team: Product & AI
Reports to: CEO
Compensation: $150K–$200K + meaningful early-stage equity + unlimited PTO

Inhouse is the #1 AI lawyer for small and midsize businesses. We combine AI, our own law firm, and an expert feedback loop to deliver fast, compliant, high-quality legal work at a fraction of the cost and turnaround time of traditional firms. We grew revenue 1,500% last year and recently raised a $5M seed round from leading VCs and the former CEO & Cofounder of LegalZoom.

You are a former practicing lawyer who fell in love with building. Over the last couple of years, you've been writing prompts, designing workflows, and shipping real things with LLMs — and you have strong opinions about what "good" looks like in legal AI.

As our Founding Legal Engineer, you will own how Inhouse's AI thinks like a lawyer. You'll design the workflows, write the prompts, build the evals, and curate the datasets that turn our product from "impressive demo" into "reliably better than a junior associate" on the legal tasks our SMB customers actually face.

Take the legal tasks our customers face — hiring employees, signing MSAs, data privacy addenda, equity arrangements, leases, terms of service — and turn them into repeatable AI workflows. Break complex legal reasoning into structured steps and decision trees that an AI can follow and a non-lawyer can understand. Define what "good" looks like for each workflow, including acceptable risk levels, fallback explanations, and escalation points.

Craft and refine the prompts and system instructions that drive drafting and review of contracts, NDAs, service agreements, employment docs, shareholder agreements, and leases. You'll own the prompt library and treat it like production code — versioned, tested, and continuously improved.

Design evaluation frameworks for AI outputs: issue spotting, risk grading, redline quality, plain-language explanations, and jurisdictional nuance. Build and curate gold-standard datasets and test suites for key use cases. Use them to benchmark model performance, catch regressions, and prove when a change actually made things better.

Partner with our attorney network to pull their playbooks out of their heads and into reusable AI behaviors and templates. Serve as the translator between how lawyers actually work and how our models need to be instructed.

Set clear boundaries for what the AI should and shouldn't do, when to recommend a human attorney, and how to communicate limitations in plain language. Make sure our flows align with Inhouse's "not a law firm" status while still delivering meaningful, actionable guidance.

As the first hire on the Product & AI team, you'll naturally be close to product decisions. Expect to:

• Join customer interviews and user research calls; bring legal insight to what you hear
• Build no-code prototypes to test new workflows before engineering invests
• Write lightweight specs for the legal AI features you're driving
• Partner with GTM and support on friction points in legal workflows
• Help shape user-facing content that demystifies legal concepts

Formal PM experience isn't required and isn't the focus.

You've shipped functional things with LLM tools — prompt libraries, workflows, agents, evals, internal tools, side projects, prototypes
You have strong opinions about what distinguishes good legal AI output from bad, and you can defend them with examples
You're comfortable with structured data, templates, APIs, and prompt formats. You don't need to write production code, but you can read it and wire things together

• 2–5+ years at a law firm or in-house legal team
• You've personally drafted, reviewed, and negotiated the kinds of agreements our users work with — MSAs, NDAs, employment and equity docs, vendor agreements

• You're comfortable running lightweight experiments and iterating on qualitative and quantitative feedback
• You enjoy simplifying complex issues into intuitive flows for non-lawyers

• Excited to join a post-seed company where you'll ship, not advise
• Biased toward action, comfortable with ambiguity, happy working with incomplete information
• You want to be measured by what you build, not what you recommend

• Prior experience in legal tech, knowledge management, or legal operations
• Experience with eval frameworks (Braintrust, LangSmith, custom) or LLM-as-a-judge setups
• Experience explaining legal concepts to non-lawyer stakeholders — founders, operators, sales, CS

Send us a resume and — idealy— one or two examples of legal AI work you've built. A prompt you're proud of, a workflow you designed, an eval set, a prototype, a Loom walkthrough, anything. ryan@inhouse.ai