RACESIDE

Historical form intelligence for UK & Irish racing. Search any trainer, jockey, owner, sire, dam, damsire or horse and pull their full historical record — strike rates, A/E and level-stake P/L — sliced by course, distance, going, class, code, region and date range.

Data: The Racing API analysis endpoints (Standard plan or above).

Stack

• index.html — the whole app, no build step
• api/tra.js — Vercel serverless proxy (holds Racing API credentials, whitelists paths, edge-caches responses for 1h)

Deploy (phone-friendly)

1. GitHub → New repository → name it raceside (private is fine) → create.
2. Add the three files (index.html, api/tra.js, README.md) via Add file → Upload files or Create new file (for api/tra.js type api/tra.js as the filename to create the folder).
3. Vercel → Add New → Project → import raceside → Framework preset: Other → Deploy.
4. Project → Settings → Environment Variables → add:
   • RACING_API_USERNAME — your Racing API username
   • RACING_API_PASSWORD — your Racing API password (same values as prizerun/racing1 — copy them across)
5. Deployments → ⋯ → Redeploy (env vars only reach functions on a fresh deploy).

Open the site, pick TRAINERS, type a name, tap FIND.

Sanity check

Visit /api/tra?path=/v1/trainers/search&name=appleby on your deployment:

• JSON with search results → working
• {"ok":false,"error":"no-credentials"} → env vars not saved or no redeploy yet
• 401/403 from upstream → wrong credentials or plan doesn't cover the endpoint

Notes

• Analysis endpoints are Standard plan. Raw per-entity /results history is Pro — the proxy already whitelists those paths, so the app can grow into them without changes.
• Edge caching (1h) keeps you well inside the API rate limit even if you hammer it.
• Distance filters are entered in furlongs and converted to yards (×220) for the API.