---
title: Use cases
---

Each page answers one question people and agents ask when they look for a tool: what Loom
does for that need and where to start.

| Need                                                                             | Page                                                       |
| -------------------------------------------------------------------------------- | ---------------------------------------------------------- |
| Memory across sessions and tools for coding agents and other AI agents           | [Agent memory](/docs/use-cases/agent-memory)                            |
| Knowledge and document search for agents (RAG)                                   | [Document search](/docs/use-cases/document-search)                      |
| A code index and code intelligence for coding agents                             | [Code index](/docs/use-cases/code-intelligence)                         |
| What the code index gives per language (C++, Python, TypeScript, Go, Rust, Java) | [Code index by language](/docs/use-cases/code-intelligence-by-language) |
| Memory linked to the code it is about, flagged when the code changes             | [Memory linked to code](/docs/use-cases/memory-linked-to-code)          |
| Lower model cost for coding agents                                               | [Coding-agent cost](/docs/use-cases/coding-agent-cost)                  |
| Software-factory memory and handoffs between implementation and review agents    | [Multi-agent](/docs/use-cases/multi-agent)                              |
| Memory, recall, audit and erasure for autonomous agents                          | [Autonomous agents](/docs/use-cases/autonomous-agents)                  |
| Agent runs, traces, tracker and PR history                                       | [Runs and PR history](/docs/use-cases/runs-and-pr-history)              |
| Self-hosted agent memory on ClickHouse, with Docker or Kubernetes                | [Self-hosted on ClickHouse](/docs/use-cases/self-hosted-clickhouse)     |

[When Loom does not fit](/docs/getting-started/overview#when-it-does-not-fit) lists the
needs another tool serves better.
