---
title: "How we’re building a data platform for a new user: agents"
description: "AI agents are emerging as autonomous database users that generate SQL, execute queries, and iterate in seconds. Because traditional analytics workflows built for human analysts cannot keep up with these rapid loops, ClickHouse is adapting its platform for agentic workloads. The proposed Agentic Data Stack combines LibreChat for conversational interfaces, ClickHouse with Model Context Protocol servers for the data layer, and Langfuse for observability. Supporting developer tooling includes chDB 4 with a Pandas-like API, IDE integrations, and ClickHouse Agent Skills to enforce database design best practices. Platforms like Anthropic and Character.AI already leverage ClickHouse's high-concurrency aggregation and real-time ingestion capabilities to handle massive data scales required by modern AI applications."
---

# How we’re building a data platform for a new user: agents

[Clickhouse](https://yomu.fyi/company/clickhouse) · Al Brown · Mar 27, 2026

**Type:** Explainer

## Summary

AI agents are emerging as autonomous database users that generate SQL, execute queries, and iterate in seconds. Because traditional analytics workflows built for human analysts cannot keep up with these rapid loops, ClickHouse is adapting its platform for agentic workloads. The proposed Agentic Data Stack combines LibreChat for conversational interfaces, ClickHouse with Model Context Protocol servers for the data layer, and Langfuse for observability. Supporting developer tooling includes chDB 4 with a Pandas-like API, IDE integrations, and ClickHouse Agent Skills to enforce database design best practices. Platforms like Anthropic and Character.AI already leverage ClickHouse's high-concurrency aggregation and real-time ingestion capabilities to handle massive data scales required by modern AI applications.

## Context

AI agents act as a new type of database consumer requiring real-time ingestion, low-latency concurrency, and fast iteration. Traditional analytics patterns involving human analysts writing SQL and building dashboards are too slow for automated workflows that reason, generate queries, and inspect schemas in seconds.

## Approach / What changed

ClickHouse introduced the Agentic Data Stack combining LibreChat, ClickHouse via MCP, and Langfuse for LLM observability. Additional tooling includes open-source and Cloud MCP servers, chDB 4 with a Pandas-like API, IDE plugins, agent-specific best-practice packages, and documentation formatted for automated agent consumption via llms.txt and Markdown.

## Takeaways

- The Agentic Data Stack integrates LibreChat, ClickHouse through Model Context Protocol servers, and Langfuse observability into a pre-wired Docker deployment.
- Internal deployment of this architecture via the DWAINE assistant handles approximately 70% of data warehouse queries and reduced analyst workload by 50% to 70%.
- chDB 4 introduces an embedded Pandas-like DataStore API featuring lazy execution, filter pushdown, and column pruning to bridge LLM code generation with the ClickHouse engine.

**Tags:** [Architecture](https://yomu.fyi/topic/architecture), [LLMs](https://yomu.fyi/topic/llm), [Observability](https://yomu.fyi/topic/observability), [Open Source](https://yomu.fyi/topic/open-source), [Scalability](https://yomu.fyi/topic/scalability)

- Source: [Clickhouse](https://clickhouse.com/blog/building-a-data-platform-for-agents)
- Source URL: https://clickhouse.com/blog/building-a-data-platform-for-agents
- Ingested by Yomu: 2026-08-28T01:23:17.956Z

[Read original post](https://clickhouse.com/blog/building-a-data-platform-for-agents)