mindsdb

<a name="readme-top"></a>

<p align="center">
  <a href="https://github.com/mindsdb/mindsdb">
    <img src="/assets/mindsdb-header-github.png" alt="Query engine for AI analytics, powering agents to answer questions across all your live data" width="100%" />
  </a>
</p>

<div align="center">
  <a href="https://pypi.org/project/MindsDB/" target="_blank">
    <img src="https://badge.fury.io/py/MindsDB.svg" alt="MindsDB Release" />
  </a>
  <a href="https://www.python.org/downloads/" target="_blank">
    <img src="https://img.shields.io/badge/python-3.10.x%7C%203.11.x%7C%203.12.x%7C%203.13.x-brightgreen.svg" alt="Python supported" />
  </a>
  <a href="https://hub.docker.com/r/mindsdb/mindsdb" target="_blank">
  <img src="https://img.shields.io/docker/pulls/mindsdb/mindsdb.svg?logo=docker&label=Docker%20pulls&cacheSeconds=86400" alt="Docker pulls" />
  </a>

  <p align="center">
    <a href="https://mindsdb.com?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">Website</a>
    ·
    <a href="https://docs.mindsdb.com?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">Docs</a>
    ·
    <a href="https://mindsdb.com/contact?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">Contact us for a demo</a>
    ·
    <a href="https://mindsdb.com/joincommunity?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">Community Slack</a>
  </p>
</div>

---

MindsDB is a popular open-source query engine for AI analytics, powering AI agents that need to answer questions directly from databases, data warehouses, and applications, with no ETL required.

<div align="center">
  <a href="https://www.youtube.com/watch?v=HN4fHtS4mvo" title="Watch demo">
    <img
      src="/assets/mindsdb_demo.gif"
      alt="MindsDB demo - answer questions in plain English from live enterprise data"
      title="Click to watch the full video"
      width="80%"
    />
  </a>
</div>

## What you can build with MindsDB Query Engine

| CONVERSATIONAL ANALYTICS AGENTS | SEMANTIC SEARCH AGENTS |
| --- | --- |
| Get precise, data-driven answers using natural language. <br /><br /> Unify and query data across sources (MySQL, Salesforce, Shopify, etc.), without ETL. <br /><br /> <a href="https://www.youtube.com/watch?v=QIdPpzcaxXg">Watch video</a> | Ground LLM responses in your most relevant internal knowledge. <br /><br /> Search across unstructured sources like documents, support tickets, Google Drive, and more. <br /><br /> <a href="https://www.youtube.com/watch?v=HN4fHtS4mvo">Watch video</a> |

## How MindsDB works

MindsDB follows a simple workflow: **Connect → Unify → Respond**. At the center is an SQL-compatible data language with additional constructs for searching unstructured data, managing workflows (jobs/triggers), and building agents.

<table style="width:100%; border-collapse:collapse; border:none;">
  <tr>
    <td style="width:25%; border:none; padding:8px 16px; vertical-align:middle;">
      <strong><a href="https://docs.mindsdb.com/integrations/data-overview?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">Connect</a></strong>
    </td>
    <td style="width:75%; border:none; padding:8px 16px; vertical-align:middle;">
      <strong>Universal data access:</strong> Give your agents federated access to 200+ live data sources (Postgres, MongoDB, Slack, files, and more).
    </td>
  </tr>
  <tr>
    <td style="width:25%; border:none; padding:8px 16px; vertical-align:middle;">
      <strong><a href="https://docs.mindsdb.com/mindsdb-unify?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">Unify</a></strong>
    </td>
    <td style="width:75%; border:none; padding:8px 16px; vertical-align:middle;">
      <strong>Dynamic context engine:</strong> Fuse structured tables with vectorized data (text, PDFs, HTML) inside a Knowledge Base.
    </td>
  </tr>
  <tr>
    <td style="width:25%; border:none; padding:8px 16px; vertical-align:middle;">
      <strong><a href="https://docs.mindsdb.com/mindsdb-respond?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">Respond</a></strong>
    </td>
    <td style="width:75%; border:none; padding:8px 16px; vertical-align:middle;">
      <strong>Autonomous reasoning:</strong> Deploy agents that blend and retrieve data points across your stack to produce grounded answers.
    </td>
  </tr>
</table>

## Setup

Users can install MindsDB via <a href="https://docs.mindsdb.com/setup/self-hosted/docker?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">Docker</a>, <a href="https://docs.mindsdb.com/setup/self-hosted/docker-desktop?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">Docker Extension</a>, or <a href="https://docs.mindsdb.com/contribute/install?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">PyPI</a>.

Here is how to pull and run MindsDB via Docker:
```bash
docker run --name mindsdb_container \
-e MINDSDB_APIS=http,mysql \
-p 47334:47334 -p 47335:47335 \
mindsdb/mindsdb:latest
```

## Usage

**Follow the <a href="https://docs.mindsdb.com/quickstart-tutorial?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">quickstart guide</a> to get started with MindsDB using our demo data.**

Retrieve and analyze data from over 200 <a href="https://docs.mindsdb.com/integrations/data-overview?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">data sources</a> in one SQL dialect. For AI agents, this means faster response time, better accuracy, and lower token consumption.
```sql
--use SQL to aggregate pipeline data from Salesforce 
SELECT SUM(ExpectedRevenue) AS open_pipeline
FROM salesforce.opportunities
WHERE close_date >= CURDATE()

--use the same dialect to retrieve even from a non-SQL database, like MondoDB
SELECT COUNT(*) AS negative_emails_last_30_days
FROM mongodb.support_tickets
WHERE sentiment = 'negative'
  AND created_at >= CURRENT_DATE - INTERVAL '30 days';
```

Create <a href="https://docs.mindsdb.com/mindsdb_sql/sql/create/view?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">views</a> and join data even from different types of data systems.
```sql
--join MongoDB and Salesforce data
CREATE VIEW risky_renewals AS (
SELECT *
FROM mongodb.support_tickets AS reviews
JOIN salesforce.opportunities AS deals
  ON reviews.customer_domain = deals.customer_domain
WHERE deals.type = "renewal"
  AND reviews.sentiment = "negative"
);
```

Join vectorized and structured data inside a <a href="https://docs.mindsdb.com/mindsdb_sql/knowledge_bases/overview?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">knowledge base</a>. Combine semantic search with precise metadata criteria in a single SQL query.
```sql
--create a knowledge base for customer issues 
CREATE KNOWLEDGE_BASE customers_issues
USING
  storage = my_vector.db,
  content_columns = ['ticket_description'];
  metadata_columns = ['customer_name', 'segment', 'revenue', 'is_pending_renewal'];

--find large customers who submitted ticket related to data security topics  
SELECT * FROM customers_issues
WHERE content = 'data security'
AND
  is_pending_renewal = 'true'.
  revenue > 1000000;
```

Use MindsDB pre-packaged <a href="https://docs.mindsdb.com/mindsdb_sql/agents/agent_syntax?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">data agents</a> and connect them with your own. See how to use MindsDB via <a href="https://docs.mindsdb.com/overview_sdks_apis?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">API or MCP</a>.
```sql
CREATE AGENT my_agent
USING
    model = {
        "provider": "openai",
        "model_name" : "gpt-xx",
        "api_key": "sk-..."
    },
    data = {
         "knowledge_bases": ["mindsdb.customer_issues"],
         "tables": ["salesforce.opportunities", "postgres.sales", "mongodb.support_tickets"]
    },
    prompt_template = 'my prompt template and agent guidance';
```
See MindsDB’s recommended usage of agents <a href="https://docs.mindsdb.com/mindsdb_sql/agents/agent?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">here</a> and how to automate workflows with <a href="https://docs.mindsdb.com/mindsdb_sql/sql/create/jobs?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">jobs</a>.

## 📃 Tutorials
- Enterprise Knowledge Search (<a href="https://mindsdb.com/blog/fast-track-knowledge-bases-how-to-build-semantic-ai-search-by-andriy-burkov?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">example</a>)
- Advanced Semantic Search  (<a href="https://mindsdb.com/blog/blend-hybrid-retrieval-with-structured-data-using-mindsdb-knowledge-bases?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">example</a>)
- Customer Support Automation (<a href="https://mindsdb.com/blog/building-janus-an-ai-powered-customer-support-helpdesk-system-powered-by-mindsdb?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">example1</a>, <a href="https://mindsdb.com/blog/building-agentic-workflow-auto-banking-customer-service-with-mindsdb?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">example2</a>)
- Intelligent Content Discovery (<a href="https://mindsdb.com/blog/mysql-mindsdb-unlocks-intelligent-content-discovery-for-web-cms-with-knowledge-bases-and-cursor?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">example</a>)
- Financial Analysis Agents (<a href="https://mindsdb.com/blog/streamline-financial-analysis-with-mindsdb-s-knowledge-bases-and-hybrid-search?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">example</a>)
- Real-time AI-powered analytics (<a href="https://mindsdb.com/blog/mariadb-mindsdb-turns-woocommerce-data-to-insights-with-real-time-ai-analytics-for-ecommerce-teams?utm_medium=community&utm_source=github&utm_campaign=mindsdb%20repo">example</a>)
- Conversational Data Assistants (<a href="https://mindsdb.com/blog/unlocking-operational-intelligence-in-energy-util