Founding Engineer

Oxus automates internal audit workflows for public and pre-IPO companies. Companies including leading payroll providers, public biotechs, and social media enterprises run their internal audits through Oxus.

Internal audit teams at some of the world's largest companies still spend enormous amounts of time manually gathering, reviewing, and documenting information to make sure their businesses are operating as they should. Oxus uses AI to automate this work while preserving the traceability, reviewability, and judgment that audit requires.

We’re a founding team of three MIT grads and are backed by Y Combinator, Paul Graham, Mathilde Collin, Depth Ventures, Robinhood Ventures, and more.

We're looking for our first Founding Engineer to join us. No audit or accounting experience is required.

• Build agentic workflows that perform complex, multi-step audit tasks
• Work on retrieval, structured extraction, citations/provenance, OCR, and long-context reasoning across complex documents and messy enterprise data
• Own and ship features end-to-end, from idea to customers
• Diagnose issues quickly and ship fixes
• Proactively identify and suggest product and technical improvements

• Week one: Ship your first feature to production
• Month one: Own increasingly complex features end-to-end
• Quarter one: Own a meaningful surface area of the product, proactively identifying improvements and driving solutions from idea to production

• A track record of building and shipping real software, whether through work, internships, startups, or ambitious projects
• Strong Python and backend engineering skills
• Significant experience building with Claude, Codex, Cursor, agent SDKs, and other AI tools
• Moves quickly while maintaining a high bar for quality, with strong judgment about when to investigate further or ask
• Clear, proactive communicator
• Comfortable figuring things out without an established playbook
• Excited to work intensely alongside the founding team: we're an in-person team in San Francisco and typically work 9/9/6

We care much more about what you've built, how you think, and how quickly and efficiently you can execute than checking specific credential boxes.

• Experience with document AI, RAG, OCR, multimodal models, or large-scale document processing
• Experience building full-stack products
• Experience with AWS, Kubernetes/EKS, and/or distributed systems
• Built your own products, startups, or technically ambitious side projects
• Been an early hire at a startup or worked on an early-stage team

You'll join early enough to have a meaningful impact on the product, architecture, and engineering culture. You'll work on real AI systems used in high-stakes workflows, where outputs need to be accurate, traceable, and reviewer-ready.

We're a close-knit team that works hard, moves quickly, and invests in making the experience fun. Perks include free lunch and dinner, unlimited coffee/matcha/boba, team outings and offsites, and the AI tools you need to do your best work :)

1. Intro call
2. Technical interview
3. In-person work trial
4. Offer + celebratory dinner with the team