Selected work
Athena Relay
Internal AI infrastructure for Business Canvas's Athena service.
What Athena is
Athena is Business Canvas’s service for helping B2B teams identify the companies and decision-makers they should meet next, using observable market and buying signals to connect research with real sales conversations.
What I built
Athena Relay is the internal system supporting that work. Rather than replacing every tool with a single interface, it connects developer tools, shared team context, and operating workflows while preserving ownership, provenance, and privacy boundaries.
Why it mattered
AI tools can make one person faster while leaving their context, decisions, and outputs scattered across separate environments. That gap makes useful work difficult to reuse and even harder to trust as a team.
My role
I work across product definition, developer-facing interfaces, client and backend contracts, security boundaries, rollout, and operational verification. The work requires treating the local developer experience and the production system as one end-to-end product.
What I learned
The quality of an AI system depends on more than model capability. Context ownership, consent, observable failure modes, and evidence that the current system actually performed the work are equally important.