Turning expert calls into data
Engineering leader at Tegus on building a data platform for expert interviews
Tegus won by turning each expert call from a one time service event into inventory that could be sold again and again. That changed the unit economics from brokered calls to software like seat subscriptions. The call itself became content creation, and the real product became a searchable library that investors could mine across companies, competitors, and themes, then connect to models and filings inside one research workflow.
-
Traditional networks like GLG made money by charging high markups on each call and treating experts as the core asset. Tegus charged calls close to cost, around $300 to $400, because the bigger prize was a roughly $25K per seat library subscription. That made reusability, not brokerage margin, the center of the business.
-
The transcript library only became valuable because Tegus added structure on top of raw text. Entity tagging, summaries, company cross links, and search let an analyst jump straight to the 30 seconds in a long call that mentioned a target company, instead of reading the whole transcript. That is what made expert calls behave like data.
-
This model pulled Tegus toward a broader research stack. Canalyst models and BamSEC filings turned transcript insight into something investors could compare against numbers and source documents. That is also why AlphaSense bought Tegus, because proprietary content libraries are harder to replicate than search interfaces alone, and competitors are now moving toward similar subscription transcript products.
The next phase is turning transcript libraries into workflow software. As AI makes summarizing and searching easier, the durable advantage shifts to owning proprietary expert content, linking it to adjacent datasets, and embedding it into the analyst's daily tools through APIs, alerts, and model workflows. That pushes the market further away from pure expert brokering and toward integrated research platforms.