Market data foundation
Normalize CSV imports, provider polling, live candles, and Go worker ingestion into one auditable Postgres model.
Final-candle analysisProvider healthGap recovery
API https://ai-trading-agent-api.vercel.app
QuantRift combines deterministic signal analysis, grounded AI research, evidence quality, and human-reviewed paper planning in one auditable platform.
Readiness
92%
Review first
07
Data health
GOOD
Deterministic engines classify. AI explains supplied evidence. Humans review the result.
No broker executionPlatform
Use the full stack or replace modules behind typed, auditable boundaries as research requirements evolve.
Normalize CSV imports, provider polling, live candles, and Go worker ingestion into one auditable Postgres model.
Build signals from typed indicators, patterns, quality gates, market regimes, evidence, and reproducible engine runs.
Compose explanations and paper-only plans from persisted evidence without letting agents rewrite source signals.
Start with Neon-backed accounts, workspace RBAC, job queues, worker health, audit trails, and fail-soft dashboards.
Architecture
Python remains the product brain. Go handles bounded ingestion work. Neon stores the auditable source of truth.
Quick start
Bring up the development stack, apply the schema, and seed deterministic defaults.
$ make dev
$ make migrate
$ make seed