FactSet as AI System of Record
SVP of Technology & Product Strategy at FactSet on driving trust through auditability
The winner in AI for investment research will be the platform that becomes the system of record for machines, not just humans. In practice that means a hedge fund, bank, or wealth manager will want an agent to pull prices, estimates, filings, notes, and internal research through APIs, then return source linked answers inside the firm’s own tools. FactSet is building toward that with Mercury, Conversational API, and packaged data feeds for AI workflows.
-
This is a shift from screen based workflow to machine readable workflow. FactSet already frames Mercury as an agentic layer, and its Conversational API lets clients embed that into their own stack instead of forcing users back into a terminal window.
-
The competitive bar is not just having data, it is exposing that data in forms agents can actually use. S&P Capital IQ also sells API access, plug ins, and bulk feeds, so the battleground becomes coverage, permissions, latency, and whether answers stay auditable down to the underlying source.
-
AI native research tools like AlphaSense attack the reading and synthesis layer, but they still sit next to the core data workstations in many accounts. That leaves incumbents like FactSet in a strong position if they can turn their structured datasets and entrenched workflows into agent ready infrastructure.
Over the next several years, investment data platforms will look more like operating systems for financial agents. FactSet’s path is to make its data, models, and internal research hooks easy to call from software, easy to audit, and hard to replace inside enterprise workflows. If it does that well, AI expands the value of the platform instead of displacing it.