AlphaSense's Bet on Expert Transcripts

This bet is really a bet that proprietary private market conversation data will matter more as public information search gets cheaper. AlphaSense can already search filings, earnings calls, broker research, and web data. The Tegus library adds information that is harder to scrape because it only exists after a real expert call. Tegus also built the operating machine to source experts fast, run calls at cost, transcribe everything, and turn each call into a reusable asset for later users.

• The model scales less like software alone and more like a factory that improves with software. Tegus had an operations team focused on discovery, outreach, scheduling, and transcription, with technology reducing the time to find experts and AI making each transcript easier to tag, summarize, and cross link after the call is done.
  1 sacra
• The economics work because the transcript library is the real product. Tegus priced calls near pass through cost, around $300 to $400 per call, while monetizing a subscription library at roughly $25K per seat. That lets one expensive human interaction generate value for many later readers instead of only the original caller.
  1 sacra 2 sacra
• Quality matters more than raw volume. Tegus differentiated by using investor generated calls and higher bar experts, while some rival libraries were described as broader but less targeted. That matters because the value of the tenth useful transcript is much higher than the value of a large pile of generic calls no one trusts enough to use in real underwriting.
  1 sacra 2 sacra 5 sacra

Going forward, the winning platforms will combine fresh expert content with AI that turns thousands of transcripts into searchable themes, recurring questions, and links to models and filings. That pushes the market toward integrated research workstations, where the moat is not just owning transcripts, but owning the full loop of sourcing, structuring, and surfacing hard to find private market knowledge.