Health, home and travel decision
Brawler
A capable developer can reproduce a useful subset (training plans, video upload + basic pose analysis, LLM coaching) but the full product (polished native app, real-time AR overlays, production-grade low-latency video inference and proprietary model improvements) is not realistic to fully replicate quickly.
View on the App Store↗Not priced
No pricing page we fetched carried a figure, so there is nothing to compare against. The build side is still real.
$100one-off52 h to build
$200/mo6 h/mo upkeep
No published price to break even against.
No open-source build does this yet
Nothing published replaces this one, so a replacement starts from an empty file. Here is what it would have to cover.
What a replacement has to do
- User records or uploads boxing video → run pose/video analysis to extract metrics → generate corrective coaching feedback and drill recommendations via an LLM/prompted model → present feedback and log workout/progress; optionally log meals via photo/barcode for nutrition tracking.
What it still won’t have
- Polished native iOS UX and App Store polish
- Proprietary model improvements and any private training data used by vendor
- Scalable, production-hardened video pipeline and low-latency mobile inference
- Built-in AR/real-time skeleton overlays and Apple Watch integration
- App store distribution, existing user base, and reputation
What remains hard
- Product polish and ongoing maintenance
First-year cost
No published price
Brawler does not publish a price we could read, so there is nothing to compare against. What building costs is below.
Money you would actually spend
Time you would spend
—
What you would spend
What we assumed
The verdict above measures whether you could build it. This one is only about money.
Runnable build prompt
Build a minimal AI boxing coaching web app and companion simple iOS client using Next.js (React) for the frontend, Node.js/Express for the API, Postgres for data, AWS S3 for media, and Stripe for subscriptions. Integrate MediaPipe (or an open-source pose detector) as a backend job to process uploaded videos and extract per-frame pose/keypoint metrics; persist results in Postgres. Use an LLM (OpenAI-compatible) to map extracted metrics to concise corrective feedback and recommended drills (prompted templates). Implement features in scope: user auth, video upload, queued video processing worker, results page showing timestamps + feedback + saved history, structured training programs (create/read), nutrition photo ingestion that extracts text/barcode and estimates calories via LLM, and basic subscription gating (Stripe). Out of scope: AR overlays, Apple Watch integration, federated/team management, and on-device low-latency inference. Require proper error handling, background job retries, API and unit tests for core endpoints, and Docker-based deployment scripts.
How we checked
How the score was reached
- Partly verdict base52
- 2 cited sources+1
- 3/3 assessment runs agreed+4
- Evidence score57
The base comes from the verdict. Everything under it is a check that either happened or did not, and each one is a fact frozen in this record rather than a judgement made at render time - so the same evidence always produces the same number.
How scoring works →Cited sources · 2
Every page the run actually retrieved.
- official productBRAWLER- Future of Boxing App - App Store
- official pricingSportsReflector: AI Coach App - App Store (pricing source provided)
Integrity checks
What held up, and what did not.
