Michael Blau is building Drip, a service that prepares financial content for AI agents. His current public profile describes earlier work at a16z; the firm also retains an author page for his published work.[2][1]
Start with Drip →
Links
X Contributions
Prepare financial content before asking an agent to use it
Blau describes a recurring problem in financial-agent applications: newsletters, podcasts, video and social posts arrive in inconsistent forms and are expensive to ingest and structure. His Drip work targets that preparation layer, so a downstream application can focus on its agent and interface.[2]
Turn commentary into structured records
Drip extracts stock mentions into records containing a ticker, sentiment, action, conviction and original source quotation. Applications can filter the resulting feed by date, analyst or ticker through a REST API, while an MCP endpoint exposes it to agent clients. The same structured data can support an interface showing agreement or disagreement among commentators without making the agent parse each original episode or newsletter anew.[3]
Tools, writing & talks
Apps
Start here
Dripdripstack.comStructures financial commentary into records with tickers, sentiment, source quotations and analyst information, accessible through APIs and MCP.
Start with the example stock-pick records and API/MCP access paths. Browsing the product explanation needs no coding; live data access requires a subscription or configured pay-per-request agent wallet, and API integration requires HTTP/authentication knowledge.
Tools
- Agentic Reservationsagentres.dev
API that lets AI agents search and book restaurant reservations on Resy using crypto micropayments.
Related projectOriginal source
Writing, research & talks
A first-party account of preparing unstructured financial content for investing and trading agents.
Builder describing the product[2]