About the role
Vantum is building the operating system for fractional operators — professionals managing 3-6 simultaneous client engagements who need a "second decision brain" to allocate their attention across competing priorities. We're pre product-market fit, running a structured 8-10 week customer development program with 8-10 committed design partners starting June. They'll be our first paying customers and give us the social proof to scale to our next 100.
We're a two-person team — Tabarak (CEO, product, customer development) and Brian (COO, design and full-stack frontend) — looking for our third core engineer. You'd own the backend and AI systems: Prisma schema and migrations, Next.js API endpoints, LLM-powered features (the agent chat, witness voice generation, step enrichment pipeline), and the prompt engineering that makes the product feel like it actually knows the operator.
Stack: Next.js 15, TypeScript, PostgreSQL, Prisma, Supabase auth, OpenRouter (primarily Claude). Hosted on Railway. We use Cursor and Claude Code heavily.
What this role is: meaningful ownership over the systems that ship to our design partners and shape what V1 becomes. Direct collaboration with the founders — you'll be in product decisions, not just executing them. The product will evolve as we learn from real users, so you'll need to be comfortable with iteration based on customer signal.
What this role is not: a stable Series A engineering role with frozen specs and quarterly planning. We're early, the bet is real, and what gets built next gets shaped by what design partners tell us during the program.
What we're looking for: a strong backend engineer who treats AI integration as a first-class skill. You've shipped LLM-powered features in production, you have opinions about when to use streaming vs structured outputs vs tool calls, you've tuned prompts after watching them fail in interesting ways, and you can write a clean API endpoint without supervision. You don't need ML research depth — you need backend craft plus AI fluency.
Bonus if you've worked at a pre-PMF startup before and know what customer development feels like in practice. We're running this deliberately, not flying blind.
You'd be working closely with Tabarak (Bangkok, GMT+7) — same timezone is ideal, +/- 3 hours is workable. Async by default but real-time sometimes when designing systems together.
We're hiring one engineer now and plan to grow the team after this hire ramps — your shape and judgment will influence who we add next.
Compensation: cash range listed below, with an opportunity for meaningful equity after a successful first year. We believe in earning ownership through tenure and contribution rather than handing it out at hire.
