November 27, 2025: Stock brokerage major Zerodha has pumped $5 million into Tijori Finance, its long-time data partner, just as the startup begins reshaping itself into an AI-first enterprise product company. The investment highlights a clear shift in the fintech-data landscape: clean, structured financial data is becoming the new infrastructure layer for AI-driven research and analytics, and Tijori wants to be the company powering it.
Tijori, well known as the data engine behind Zerodha’s Kite research features, is now widening its lens beyond retail users. Co-founder Siddharth Hegde said the company is transitioning its revenue engine toward enterprises and global markets, using AI as the multiplier.
“We want to take all the clean data we’ve built and turn it into micro AI products that solve very specific pain points,” Hegde stated. “We’re not building one master product. We’re slicing AI into modules.”
Historically, about 80–90% of AI-first enterprise product company paid users arrived through Zerodha integrations, overwhelmingly retail investors. But that funnel is no longer the core focus.
As the product suite matures, enterprise clients, analysts, funds, research desks, will become the primary revenue driver.
With 14,000 Paid Users, Tijori Expands AI Suite Beyond Retail
“Retail users won’t pay for a 15-year management consistency check, but analysts will,” Hegde said, noting that 70% of revenue from its AI tools will come from enterprise clients.
AI-first enterprise product company currently counts 14,000 paying customers, alongside revenue from its Kite integration.
Much of Tijori’s confidence comes from its proprietary Atlas database, comprising 15 years of structured and unstructured corporate filings converted into machine-readable formats.
“Clean data is the foundation. If the underlying data is noisy, AI will give rubbish outputs,” Hegde said.
Tijori’s AI expansion started with Tijori Alerts, a system that scans nearly 4,000 daily exchange filings and pushes WhatsApp and email alerts—often within seconds—sometimes outpacing media reports.
Its flagship tool, Concall Monitor, joins live analyst calls, transcribes them in real time, and instantly generates dashboards, summaries, and consistency checks. It can compare comments on the call with 15 years of past disclosures.
“We can take Ola’s November call and match it against everything they’ve said for the last 15 years and flag any conflict,” Hegde said. “That’s 100% AI. No human intervention.”
Tijori uses a blend of AI models, Kimmy2 for about 45% of tasks, Gemini for 40%, and Claude for heavier workloads. Hegde called Gemini “the best value,” pointing out that AI-first enterprise product company now runs about 7 billion tokens a month across its system.
Tijori’s fresh capital will go directly into expanding AI capacity.
“A lot of the new funding goes straight into LLM queries, GPUs and people,” Hegde said. “The product pipeline is ready; now we need infrastructure and hires.”
AI-first enterprise product company has already ingested filings for the top 1,000 US companies and sees global markets as the next frontier.
“If we can read an Indian annual report, we can read a 10-K,” Hegde said. “We could power hedge funds or brokers in the US.”
The timing aligns with the broking firm’s upcoming US trading rollout on Kite, where Tijori will likely support research capabilities.
Despite the expansion, Tijori isn’t building a monolithic subscription like Bloomberg Terminal.
“Users get scared of one giant product,” Hegde added. “We’d rather unbundle, micro features that solve problems they didn’t even know could be automated.”
His roadmap is straightforward “We’re going all in on AI. The stack leans toward enterprises. And once India is stable, we’ll go global.”
