⛳ AI Golf Coach Case Study
From raw data to one clear action.
A deterministic golf coaching system that turns performance metrics into one focused next step instead of overwhelming players with numbers and competing advice.
Executive Snapshot
Problem: Golf launch monitors provide valuable performance data, but players can still be left unsure what to focus on next.
Goal: Turn raw performance metrics into one clear coaching action per session.
Scope: Coach · Analysis · Results · Session Goal · Metric confidence states.
Approach: Define metric logic → map confidence states → design focused coaching flows → test engine logic.
Outcome: Stable Engine v1 baseline with one focus metric per session, evidence based confidence and consistent status logic across screens.
⏱️ 2026 | 📱AI Golf Coach MVP | 🧑🏼💻 Role: Product Design, UX Flow, System Logic | ✏️ Tools: Lovable, Figma, AI assisted prototyping
🚀 See it in Action
Explore the interactive walkthrough or open the live MVP.
🧭 Strategic Narrative
More data does not always mean a clearer decision.
Golf launch monitors give players useful performance metrics, but the next step can still feel unclear. A player may see Smash Factor, Attack Angle, Launch Angle, Spin Rate and Dispersion, but still not know what to focus on first.
AI Golf Coach was designed around a simple product idea: one session should have one clear focus.
The system looks at the available metrics, evaluates the confidence behind them and turns the result into one practical coaching action. If the evidence is limited, the feedback stays broad. If the data supports a specific pattern, the coaching becomes more precise.
The system prioritises one focus area, checks the quality of the supporting data and decides how specific the coaching recommendation should be.
⚡System Level Decisions
| System Layer | Before | After |
|---|---|---|
| Metric Feedback | Too many metrics competing | One focused priority |
| Coaching Logic | Generic advice | Deterministic recommendations |
| Confidence | Same tone for all data | Feedback scaled by evidence |
| Screen Consistency | Risk of status mismatch | One source of truth |
🧠 Core Product Logic
AI Golf Coach was designed to translate golf performance data into one useful coaching action.
Instead of treating every metric as equally important, the system evaluates three core performance areas: Smash Factor, Attack Angle and Dispersion. Each metric has its own threshold logic, supporting data and confidence state.
The goal was to make the coaching useful without pretending to know more than the data supports.
Key Product Decisions
One focus metric per session
The player should not receive competing advice during one training flow.
Confidence scaled by evidence
The system only becomes precise when supporting data justifies it.
Short action based coaching
Each recommendation is written as something the player can immediately try.
One source of truth
Coach, Results and Analysis share the same status logic.
What Stayed Intentionally Simple
Manual input first
The MVP uses manual metric input so the decision logic could be tested before adding screenshot recognition.
Three core metrics only
Engine v1 focuses on Smash Factor, Attack Angle and Dispersion instead of expanding too early.
Deterministic logic
The system avoids vague AI advice and uses clear rules for each coaching state.
📊 Outcome
Engine v1 established a stable coaching baseline.
The system now supports one focus metric per session, confidence scaled by evidence and consistent logic across Coach, Results and Analysis.
What this proved
The product can turn raw golf data into one clear coaching action
The confidence model prevents fake certainty
The same engine logic can support Coach, Results and Analysis
The MVP is ready to be extended with screenshot recognition and session history
🚀 Try the System
Engine v1 established a stable coaching baseline.
The system now supports one focus metric per session, confidence scaled by evidence and consistent logic across Coach, Results and Analysis.