# obsei

> obsei is an open-source MCP server for analytics and data. Self-hosted customer feedback search, stats and themes, with PII redacted at ingest. It runs locally on your machine and is started by your AI client.

- **Kind:** MCP server
- **Pricing:** Open source
- **Platforms:** self-hosted, CLI, Linux, macOS
- **Website:** https://obsei.com
- **Repository:** https://github.com/obsei/obsei/tree/HEAD/packages/obsei
- **MCP Registry name:** io.github.obsei/obsei
- **Categories:** analytics and data, security
- **Listing updated:** 2026-11-15

## About

obsei is a Voice of Customer pipeline that pulls in reviews, tickets, calls and surveys, strips personal data before any model sees it, and stores the result encrypted on your own infrastructure. Its pitch is as a self-hosted, open-source alternative to Enterpret and Unwrap.

The MCP server lets an assistant search feedback, read stats and see themes, with every result carrying record ids it can cite rather than author identities. Customer text is flagged as untrusted, and the single write action, a proposed issue, waits for human approval. Themes built from fewer than k people are hidden, including in the MCP server.

Sources include App Store, Google Play, Zendesk, Intercom, Gong, GitHub, Bluesky, SQL and webhooks. Classification can use a local multilingual model on CPU or any OpenAI-compatible LLM, and the same issue raised in several languages is grouped into one theme. It needs Python 3.11 or newer and is licensed Apache-2.0 with no telemetry.

## Features

- **PII redaction at ingest:** Emails, phones, cards, IBANs and national IDs are removed before storage; names are optional.
- **Cited answers:** Results carry record ids an agent can cite, not author identities.
- **Cross-language themes:** The same issue in English, Japanese or Portuguese becomes one theme.
- **Routing on confidence:** Labels route to Jira, Slack, billing or a human when the model is unsure.
- **Erasure and audit:** Forget by author, source or age, export data, and keep an audit log.
- **Plugin sources:** Add a source, enricher or sink as a small Python package.

## Good for

- Ask why users in one market are unhappy this week
- Track spikes and new complaint themes with alerts
- Keep support tickets private while still analysing them
- Send confident bug reports to Jira and unsure ones to a person

## Install

### Claude Code (Terminal)

```sh
claude mcp add obsei --env OBSEI_DB_KEY='<OBSEI_DB_KEY>' -- uvx obsei
```

### Claude Desktop (claude_desktop_config.json)

```json
{
  "mcpServers": {
    "obsei": {
      "command": "uvx",
      "args": [
        "obsei"
      ],
      "env": {
        "OBSEI_DB_KEY": "<OBSEI_DB_KEY>"
      }
    }
  }
}
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "obsei": {
      "command": "uvx",
      "args": [
        "obsei"
      ],
      "env": {
        "OBSEI_DB_KEY": "<OBSEI_DB_KEY>"
      }
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "obsei": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "obsei"
      ],
      "env": {
        "OBSEI_DB_KEY": "<OBSEI_DB_KEY>"
      }
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "obsei": {
      "command": "uvx",
      "args": [
        "obsei"
      ],
      "env": {
        "OBSEI_DB_KEY": "<OBSEI_DB_KEY>"
      }
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "obsei": {
      "command": "uvx",
      "args": [
        "obsei"
      ],
      "env": {
        "OBSEI_DB_KEY": "<OBSEI_DB_KEY>"
      }
    }
  }
}
```

## FAQ

### Is it open source?

Yes, Apache-2.0, with no open-core tier according to its site.

### Does customer data leave my servers?

The site says nothing leaves your infrastructure unless you allow it, and there is zero telemetry.

### What do I need to run it?

Python 3.11 or newer; install with uv or pip and run obsei init. A container image is also published.

### Can it write to my tracker?

The only write is a proposed issue, which waits for your approval.

### Is obsei free?

Yes. obsei is open source, so you can use and self-host it for free.

### How do I add obsei to Claude Code?

Run this command in your terminal: claude mcp add obsei --env OBSEI_DB_KEY='<OBSEI_DB_KEY>' -- uvx obsei

### What are the best alternatives to obsei?

The closest alternatives to obsei are REA, Agent-Native Analytics, Google Ads + Meta Ads + SEO/GEO MCP, Tapetide - Indian Stock Market MCP and Agent Roast.

## Alternatives

- [REA](https://ai-powered.com/mcp/rea): REA is an open-source MCP server for developer tools.
- [Agent-Native Analytics](https://ai-powered.com/mcp/agent-native-analytics): Agent-Native Analytics is an open-source MCP server for analytics and data.
- [Google Ads + Meta Ads + SEO/GEO MCP](https://ai-powered.com/mcp/google-ads-meta-ads-seo-geo-mcp): Google Ads + Meta Ads + SEO/GEO MCP is a freemium MCP server for marketing and seo.
- [Tapetide - Indian Stock Market MCP](https://ai-powered.com/mcp/tapetide-indian-stock-market-mcp): Tapetide - Indian Stock Market MCP is a freemium MCP server for finance and payments.
- [Agent Roast](https://ai-powered.com/mcp/agent-roast): Agent Roast is an open-source MCP server for developer tools.
- [Midplane](https://ai-powered.com/mcp/midplane): Midplane is an open-source MCP server for databases.
- [World Monitor](https://ai-powered.com/mcp/world-monitor): World Monitor is a freemium MCP server for analytics and data.
- [CompanyData](https://ai-powered.com/mcp/companydata): CompanyData is a paid MCP server for analytics and data.

---
Source: https://ai-powered.com/mcp/obsei
