While everybody is busy stuffing an AI chatbot into every product, the folks at Tijori are using AI to build high-quality market research tools. They’re calling it Tijori Stack, and what they are doing is taking the messy and scattered universe of company disclosures, financials, filings, concalls, and regulatory documents, and extracting useful insights.

They have several products:

1. Concall Monitor Tracking concalls is annoying because the transcript comes after a few days. Then there’s the fact that you have to sit through one hour of management-speak just to find the three useful things they said. Tijori is trying to fix that.

With Concall Monitor, a few minutes after a concall ends, you get the full transcript and an AI-generated summary that extracts the key insights like pricing, margins, growth, demand, guidance, capex, risks, etc.

But the interesting thing is not just summarization but the management consistency check. Tijori tracks what management has said historically vs. what they are saying now. If a company promised something four quarters ago, the system can help track whether they followed through, changed the story, postponed the target, or pretended the earlier comment never happened.

2. Report on Demand Tijori ingests pretty much all the filings and disclosures that a company makes, from financials, concalls, exchange filings, and regulatory disclosures to other company-specific documents. Using this, they generate different kinds of company reports:

Risk Probe Report gives you all the red flags and key risks.

Management Credibility Report looks at what management said versus what they actually did.

Five-year revenue and EBITDA estimate is an AI-assisted financial forecast grounded in company data.

All these reports are useful when you are doing first-pass research on a company. If you are tracking five companies and want to quickly figure out which are bad and which are worth researching further, these reports help. Instead of reading annual reports, concalls, exchange filings, presentations, and financials from scratch, you can use these reports to build a baseline view in minutes.

3. Radar Radar is probably my favourite idea in the stack.

It allows you to define a metric or risk you care about for a particular company, and Tijori keeps scanning company disclosures and financials for relevant mentions. When the metric you define shows up, you get an alert.

And the metric does not have to be a typical financial metric.

You can track things like client concentration, employee growth, dollar revenue exposure, margin pressure, working capital stress, promoter pledging, regulatory risk, capex delays, inventory build-up, receivables, raw material pressure, or anything else that matters to your thesis. This is powerful because every investor has a different question.

Radar turns those questions into live monitors. I don’t think there are many products in India, or even in the world, that do this well.

4. Atlas When you use ChatGPT or Claude to ask questions about a listed company, the answer is based on info that scraped from the public web. This is problematic because public web data about companies is messy.

There is news, commentary, rumours, stale articles, bad research takes, and gossip. Sometimes it is useful, but often it is noise. What makes Atlas different is that it gives you answers that are grounded in company-specific material like filings, disclosures, financials, concalls, and other company data.

Instead of asking a generic chatbot, “Tell me about this company,” and getting an unreliable answer scraped from the internet, you are asking a system that is constrained by the company’s own disclosures and financial history. This ensures there are no hallucinations.

Check out these features here.

Go to link →