STAR - Senior AI Product Architect / Fractional CTO

**Title:** Fractional CTO / Principal Architect — agentic B2B SaaS **The product:** A B2B marketing platform serving large global marketing agencies and brands. We're moving from a workflow UI to an agent-first control plane — where AI agents act on behalf of users to run campaigns end to end. **The hard problems you'd own:** - Multi-tenant data isolation across agency offices, geographies, and competing brands — where a leak between clients is relationship-ending - A clean engagement model spanning managed-service, agency-partnership, and self-serve, without the architecture special-casing each one - A permissions layer that lets AI agents act as scoped principals — never exceeding the authority of the human they act for - Client-facing surfaces that expose exactly what's intended and nothing more **The role:** Own the technical architecture and direct a dev team building it. This is judgment and leadership, not hands-on implementation - but I want someone who could build it themselves, because that's the only kind of architect whose direction I trust. I want someone who'll push on them, not nod along. **Stack:** Node/TypeScript, Postgres, MongoDB, Clikhouse, kafka, MQ, MCP for the agent layer. Remote, fractional, ongoing. Senior rates expected. **To apply - skip the template, answer these:** 1. The hardest multi-tenant isolation problem you've owned, and the call you'd defend 2. How you'd think about letting an AI agent act for a user without it ever exceeding that user's access 3. How you've directed devs you didn't personally hire — keeping quality without becoming the bottleneck 4. Fractional availability + rate