Agent Is the New Database User: Build Agent-Facing Analytics Infrastructure

AI agents are becoming the primary users of analytics databases. This shift changes the infrastructure requirement. Agent-facing analytics needs high concurrency, sub-second response, fresh data, and an agent-friendly way to operate the database.

A human analyst submits queries one at a time, reads the results, and decides what to ask next. An agent can break one request into dozens of SQL queries, run them in parallel, inspect the answers, and immediately query again. And most of the agent's queries require filters, time windows, joins, and aggregations. The database becomes the agent's fact entry point, and database latency becomes answer latency.

In this webinar, we will look at what agent workloads demand from the database layer, how VeloDB supports real-time search and analytics over continuously updating data, and how VeloCLI plus VeloDB Skills help agents access, inspect, and operate VeloDB. VeloDB is the commercial product built on Apache Doris, an open-source real-time analytics and search database.

What you'll take away:

- CLI + Skills to connect agents to the database: Agents can query, inspect, and operate VeloDB within minutes of installing VeloCLI and VeloDB Skills. Both are open source.

- Demo: Use agents to build live dashboards with VeloDB data, and see what AI-native database capabilities can look like for agent actions and context engineering.

- Agent-facing analytics architecture: Support real-time ingestion, sub-second queries, and unified search and analytics for agent workloads.

Speakers:

Kevin Shen, Principal Product Manager, VeloDB