About the Company
L7V is a Tokyo-based venture studio that builds, backs, and operates companies end-to-end. You'll join early on an internal product: an AI-powered zero-to-launch GTM system used by the studio's own startups and eventually sold to other founders.
About the Role
The product is a web dashboard plus an MCP server, currently an MVP (TypeScript/JavaScript backend, prompt-driven core, not yet live with customers). You'll harden it into something production-ready — auth, dashboard, backend — while building out AI agents that replicate specific growth-expert workflows (starting with a paid-advertising specialist the founder works with directly).
What you'll do
- Build and maintain the web dashboard and MCP server: user auth, login flows, backend architecture, performance and reliability
- Develop AI agents and automations that encode expert GTM/growth workflows, using external AI APIs
- Design, test, and benchmark prompts and agent pipelines against real output quality
- Ship fast pre-launch (3-4 weeks), then move to a weekly improvement cadence post-launch based on customer feedback
Must-haves
- Strong JavaScript/TypeScript full-stack experience (some backend work may touch Python)
- Direct experience building AI agents/automations and integrating external AI APIs — not just prompting an existing model
- Experience building dashboards and user authentication systems
- Comfortable working from an ambiguous, evolving spec and shipping fast
Nice-to-have
- MCP server development experience
- Familiarity with agent skill frameworks for Claude Code / Cursor / Codex, or go-to-market/marketing automation systems
- Web server management and performance optimization
Logistics
- 10-30 hrs/week to start, flexible
- Timezone: min 1 hr overlap with 1-5pm JST, 2+ preferred
- Path to full-time / founding engineer role across future L7V companies for the right person