Case study
RepNotes AI
An AI-native iOS workout tracker that turns free-form workout notes into structured logs, progress analytics, insights, and next-workout plans.
ROLE
Founder / AI Product Engineer
STACK
PROBLEM
You're between sets with 60 seconds on the clock. Every workout app wants you to tap through menus and dropdowns to log what you just did. Notes apps are fast, but they leave you with a pile of unstructured text that can't track progress.
APPROACH
Log your workout the way you'd text a friend. Parser v2, an AI-assisted understanding pipeline, normalizes messy natural-language notes into exercises, sets, reps, weights, and cardio — and Plan AI turns that history into grounded insights and next-workout plans.
DEEP DIVE
RepNotes AI is an AI-native iOS workout tracker built around one idea: logging a lift shouldn't take longer than doing one. You write your workout as free-form notes, and the app turns them into structured logs, progress analytics, insights, and next-workout plans.
Parser v2 is the workout understanding pipeline. It normalizes messy natural-language notes into exercises, sets, reps, weights, cardio, notes, and training facts, and supports correction replay — so when a user fixes a parse, the correction feeds back into how their future notes are understood.
Plan AI is the coaching layer. It's facts-grounded by construction: LLM planner and narrator stages operate behind strict JSON contracts, and deterministic evidence validators check every claim against the user's actual training history — rejecting recommendations that are unsupported, stale, or fabricated.
Model behavior is tested with replayable AI evals and privacy-safe diagnostics across the parser, insights, and planning workflows — covering raw-text leakage, stale data, unsupported claims, and evidence drift. v1.0 is live on the App Store.
HIGHLIGHTS
Parser v2: normalizes messy natural-language notes into exercises, sets, reps, weights, cardio, and training facts — with correction replay
Plan AI: facts-grounded coaching built on LLM planner/narrator stages, strict JSON contracts, and deterministic evidence validators
Rejects unsupported, stale, or fabricated recommendations before they ever reach the user
Replayable AI evals and privacy-safe diagnostics covering raw-text leakage, stale data, unsupported claims, and evidence drift
Founder-built: product, data model, AI pipeline, and App Store release shipped solo