Zerodha backs Tijori Finance with $5M to scale AI-driven market research
Zerodha has invested $5 million in Bengaluru-based Tijori Finance, a market data and research platform building AI tools on top of long-term structured and unstructured datasets covering thousands of Indian companies.
Tijori started with standard exchange feeds, then expanded into parsing annual reports, investor presentations, and conference calls at scale. The thesis is simple: balance sheets and P&L are only half the picture; the rest sits in messy PDFs and inconsistent formats that most teams can't process fast enough.
This work led to a deep product integration with Zerodha. According to Tijori, Zerodha's fundamental and mutual fund datasets are powered by its pipelines today.
What finance teams can do with this
- Speed to signal: Tijori Alerts scans roughly 4,000 exchange filings daily and pushes summaries of material events (acquisitions, rating changes, leadership exits) on WhatsApp within seconds.
- Better small-cap coverage: For mid- and micro-caps where news lags, near-instant, filing-led updates narrow the information gap.
- Earnings call workflow: Concall Monitor generates full transcripts and AI summaries within a minute of the call ending, with management "consistency checks" across past calls.
- Data quality first: Clean, structured source documents reduce hallucinations and improve summary accuracy-critical if you're pushing insights to PMs or clients.
- Enterprise fit: Advanced modules are expected to be enterprise-focused, reflecting compute costs and the need for SLAs, audit trails, and access controls.
For context, these alerts typically originate from exchange disclosures. If your team relies on manual checks, compare latency against the official corporate announcements feed on the exchanges (for example, BSE corporate announcements).
Inside Tijori Stack
Tijori Stack sits on 15 years of historical filings and investor presentations. The platform focuses on cleaning and structuring complex documents so AI systems can generate context that analysts can trust.
The Concall Monitor is the standout: transcript creation, near-real-time summaries, and cross-call consistency checks that flag shifts in tone or guidance. For teams tracking dozens of names, this shaves hours off each earnings cycle.
Practical next steps for your team
- Event monitoring: Map Tijori Alerts to your watchlist and compliance flags (M&A, pledges, resignations, rating actions). Track latency and false positives for a month.
- Earnings process: Pilot Concall Monitor on upcoming calls; compare outputs with analyst notes to benchmark accuracy and coverage.
- Data diligence: Ask for documentation on data lineage, versioning of filings, deduping logic, and handling of amended disclosures.
- Enterprise guardrails: Define user-level permissions, retention, and audit logs. Set SLAs for uptime, throughput, and message delivery on WhatsApp.
- Integration path: If you use Zerodha's datasets, evaluate overlap with internal models and downstream systems before full rollout.
The signal here is clear: clean primary data plus fast summarization wins. AI adds value only when the underlying documents are reliable and searchable at scale.
If you're also reviewing your team's AI tool stack, this curated list can help benchmark options by use case: AI tools for finance.
Your membership also unlocks:
