Fit2AI 2.1: How COROS Data Enters AI Analysis
Turn COROS training history into context for conversation, moving from individual records to ongoing questions and analysis.
Insights and notes.
Thoughts on AI, software engineering and product experiences, drawn from everyday observation and practice.
From FIT Files to DaVinci Resolve: How We Build Sports Data Overlays
From parsing FIT data and synchronizing timelines to designing gauges: how workout data becomes part of DaVinci Resolve.
Why We’re Building Fit2AI
From recording training to understanding it: turning workout data into context for a conversation with AI.
From multimodal AI to agents: a turning point for applications
Multimodal capabilities and agent frameworks are taking AI applications from answering questions to understanding context, breaking down tasks and completing work.
Designing scalable event-driven architecture
Key considerations for event-driven architecture, from event modeling, idempotency and retries to observability and evolution.
Creator Widget 2.0: motion and widget updates
Refining motion, widget hierarchy and data presentation around creators’ daily routines, making growth information calmer and clearer.