Full-Stack Engineering Intern

We’re hiring a full-stack engineering intern to help build and ship a live web app used by startup founders and their investors.

The app brings a startup’s financial data into one place and connects it to the tools founders already use:

• Where data comes from: it pulls numbers from Xero and Google Sheets and keeps them in sync automatically.
• Where data lives: one PostgreSQL database, with each company’s data kept separate.
• How people use it: founders work in the web app or ask Claude, which connects through our MCP server.
• Who sees what: investors a founder invites get read-only access in Claude, limited to what the founder shares.
• AI: built-in AI features run on Google’s Gemini.

The product is already running in production, so you’ll start on real code, not a blank project.

The app is three TypeScript services sharing one PostgreSQL database, and you’ll work across all three.

• Web app (React): build and polish the screens founders use to view and manage their data.
• Backend (Node.js, Express): add API routes, data models and database migrations, and keep the Xero and Google Sheets connections reliable.
• MCP server: build the tools Claude uses to read and update a company’s data, with investors limited to what founders share.
• Integrations: Xero and Google Cloud (Sheets, Drive and Docs APIs, Google sign-in), plus AI features built on Gemini and Claude.
• Deployment: move the app to AWS and run it there, monitor the background refresh jobs, and fix what breaks in production.
• Quality and security: write tests, set up automatic test runs (CI) on GitHub, and help prepare for client security reviews.

• Solid TypeScript or JavaScript, with React on the front end and Node.js on the back end.
• SQL and a relational database, ideally PostgreSQL.
• Everyday Git and GitHub: branches, pull requests and code review.
• Hands‑on AWS: you’ve deployed an app there and kept it running, using services like EC2, ECS or App Runner, RDS for PostgreSQL, and IAM.
• Google Cloud and OAuth 2.0: enabling APIs, setting up consent screens and credentials, and building sign-in and account-connection flows.
• You can open an unfamiliar codebase of 60,000+ lines and explain how its parts fit together within your first week.
• No finance background needed, but you must be keen to learn what a P&L, a forecast and a cap table are.

• Prisma, Vite or Express.
• Building with LLM APIs (Gemini, Claude or OpenAI), function calling, or the Model Context Protocol (MCP).
• Third-party APIs such as Xero, Google Workspace or Stripe.
• Background job queues and scheduled tasks.
• Setting up CI with GitHub Actions.
• Coursework or interest in finance, accounting or startups.