Endpoints

GET /health Returns current Snowflake connection info

GET /feed?user_id=u_demo_min&limit=5 Latest transactions for user

GET /stats/category?user_id=u_demo_min&days=30 Spending stats per category

GET /predictions?user_id=u_demo_min Prediction results from ML pipeline

POST /transactions Upsert a transaction

POST /reply Upsert a user reply label (“need”/“want”)

curl -s http://127.0.0.1:8000/health | jq

{
  "U": "NGSTEPHEN1",
  "R": "ACCOUNTADMIN",
  "W": "COMPUTE_WH",
  "D": "SNOWFLAKE_LEARNING_DB",
  "S": "BALANCEIQ_CORE"
}
Feed
curl -s "http://127.0.0.1:8000/feed?user_id=u_demo_min&limit=5" | jq

Category Stats
curl -s "http://127.0.0.1:8000/stats/category?user_id=u_demo_min&days=30" | jq

Predictions
curl -s "http://127.0.0.1:8000/predictions?user_id=u_demo_min" | jq

Insert / Update a Transaction
curl -s -X POST http://127.0.0.1:8000/transactions \
  -H "content-type: application/json" \
  -d @samples/txn_sample.json | jq

{
  "id": "t_demo",
  "user_id": "u_demo_min",
  "transaction_id": "ext-demo",
  "merchant": "Starbucks",
  "amount_cents": 525,
  "currency": "USD",
  "category": "Coffee",
  "need_or_want": "want",
  "confidence": 0.9,
  "occurred_at": "2025-11-07T13:05:00-08:00"
}

checklist

•	Tables exist in SNOWFLAKE_LEARNING_DB.BALANCEIQ_CORE
•	Seed data loaded
•	/feed, /stats/category, /predictions return JSON
•	/transactions and /reply POST work
•	.env.example + README included
•	Sample JSON outputs saved under samples/