Want to know if this job is worth applying to?
San Francisco, California, United States
Onsite
Unusual Ventures is an early-stage venture firm that partners with exceptional enterprise software founders from pre-seed through Series B. We work hands-on with founders on go-to-market, customer discovery, and recruiting, with one goal: helping companies find product-market fit.
We want AI built into how we source, evaluate, support, and run the firm — not bolted on.
We're hiring our first AI Engineer to build the internal systems that can help our small team operate with scalable leverage. You'll work across the firm — investing, talent, platform, founder services, finance and operations — finding the highest-leverage problems, then designing, building, and shipping tools people use every day.
You'll be our first engineer. That means real ownership and a lot of trust, and it also means you'll set the technical standards, choose the stack, and decide what's worth building.
This is a builder role. We care more about solving the underlying problem well than wiring together the newest APIs. You'll have real ownership, direct access to partners and operators, and a short path from idea to production.
From time to time you'll also work directly with portfolio companies: helping founders scope an AI feature, pressure-test their technical approach, or stand up an internal workflow.
Talent intelligence. Agents and pipelines that find, enrich, and rank engineering and GTM talent for our portfolio — working closely with our Talent Partner to turn our recruiting judgment into repeatable systems.
Sourcing and research. Tools that surface promising companies and founders early, and context-aware research assistants that draw on our CRM, notes, meeting history, and market data to prep partners before a first meeting.
Diligence support. Workflows that speed up market maps, competitive analysis, customer-call synthesis, and memo drafting — without losing rigor.
Firm knowledge. Making what the firm already knows (thousands of founder and customer conversations, past memos, playbooks) searchable and useful.
Operations automation. Removing manual work from portfolio reporting, LP updates, and back-office processes, in partnership with the CFO/O and VP Finance.
Portfolio support. Reusable AI playbooks and tools that our founder services team can deploy across portfolio companies.
3–6 years of software engineering experience, with at least 1–2 years shipping LLM-powered products or internal tools to production
BS or MS in Computer Science, Engineering, Mathematics, or a related technical field, or equivalent practical experience
Strong Python (TypeScript a plus); comfortable across the stack, from data pipelines and APIs to a simple, usable front end
Hands-on experience with LLM APIs, retrieval, agents/tool use, and evaluation — and a clear view on when each is the wrong tool
Product taste: you talk to users, find the real problem, ship something small, and iterate
Experience working with messy real-world data (CRMs, email, calendars, third-party data providers)
Clear communicator who can explain tradeoffs to non-technical partners and earn their trust
Self-directed: comfortable setting priorities with light direction
Experience building internal tools or intelligence products at a VC firm, investment firm, or data-heavy operations team
Early-stage startup experience, or time as a founder
Familiarity with venture or enterprise software GTM
Experience with recruiting or people data
Public work — open-source projects, side projects, writing — that shows how you think and build
Small team, high leverage. Your tools will be used daily by the whole firm, and you'll see the impact directly.
Build, don't just integrate. Buy off the shelf when it's the right call; build when the problem calls for it.
In person. We work together in the Bay Area office. Being close to the people using your tools is part of the job.
Founders first. Everything we build should ultimately make us more useful to the founders we back.
Send your resume or LinkedIn, plus a short note about something you've built with AI that people actually use — what the problem was, what you shipped, and what you'd do differently. Links to code or demos are welcome.