Step 01
Step 1
Market data pulled from NSE, AMFI, RBI, fuel feeds — cached with per-type TTLs.
AI InsightsRetail + B2B
News, market data & RBI knowledge.
Live market data, financial news, and RBI / SEBI circulars — all cited, all dated.
The problem
Why this exists.
An advisor that cites a stale rate or an outdated RBI guideline is worse than one that admits it does not know.
We want a single freshness-scored knowledge surface so every answer Mukodi gives can point at the document behind it.
How it works · 3 steps
From input to outcome.
- Step 02
Step 2
News classified by impact, deduped by Jaccard similarity, scored for relevance.
- Step 03
Step 3
Regulator circulars chunked, embedded, and freshness-scored — stale circulars get superseded automatically.
Under the hood
The models and routes behind it.
Models called
API routes
- GET /api/v1/news
- GET /api/v1/market
- GET /api/v1/knowledge
Compliance posture
- Every citation carries the source URL and the date we read it.
- Stale-circular detection retires superseded documents so answers do not silently drift.
- Public knowledge routes return no per-user data and can be queried without an account.
What it's not
News and market data are informational, not investment advice. Mukodi does not recommend specific securities.
Related · features & stories