Christopher McKean Scott

Built yardr, an agentic orchestration system for routing units of work (“beads”) to whoever can execute them—across agents, scripts, people, or other machines. Designed composable concepts (beads, stages, routes, groups) so the same workflow can be executed by AI agents and humans using role-defined configuration. Implemented an event-first architecture using an append-only embedded store (bbolt) to move work forward on events rather than polling. Implemented the Read profile of the Bead Protocol (BDP) in Go with conformance testing, and developed a yardr-in-its-own-yard workflow where agents write most of the code while I set direction and review. Shipped signed release binaries for macOS and Linux and provided live visibility via a monitoring site.

Served as CTO emeritus and technical advisor starting after the main co-founding role, supporting ongoing technical leadership and guidance.

Founded PlanTime, an AI-supported scheduling platform that replaces spreadsheets and legacy tooling with adaptive scheduling workflows. Led technical discovery with early customers to find high-value scheduling use cases, scope pilot deployments, and translate operational constraints into implementation plans. Designed and shipped a production MCP server that exposes scheduling workflows (shift creation, auto-assignment, publishing) as agent tools, including OAuth-based authentication and org-scoped API keys. Developed AI-assisted and ML-inspired scheduling concepts for filling open shifts while respecting availability, and built a hierarchical multi-agent workflow (orchestrator delegating to executor agents with human review and code ownership) implemented first on Gas City and then on yardr.