Sign in
Your Saved List Become a Channel Partner Sell in AWS Marketplace Amazon Web Services Home Help

MindsDB in-database machine learning

MindsDB in-database machine learning

By: MindsDB Latest Version:

Product Overview

MindsDB, the world's fastest-growing open-source applied machine learning platform, allows developers and data engineers to quickly and effectively build AI-powered solutions using their existing skills.

Enable state-of-the-art Machine Learning frameworks as a native component of the data layer (database, data lake, data stream, etc). Train ML models as easily as running a single SQL command, and make predictions with the same experience as querying data from a table.

You can choose from 70+ native integrations with the most popular data platforms and ML frameworks, bring your custom models or even build your own integration.

Supported machine learning tasks:
Time-series (including multivariate models),
Natural Language Processing.

Min Prerequisites: SQL skills, Basic understanding of machine learning concepts.

Connect your database to the MindsDB instance using one SQL command, for example:

CREATE DATABASE your_db_name
WITH ENGINE = "db_type",
"user": "username",
"password": "password",
"host": "",
"port": "5432",
"database": "demo"

Train an ML model from your data (note that columns equate to model features) using a command like the following:

FROM your_db_name
(SELECT * FROM your_training_table_name)
PREDICT predicted_column_name;

Check the for extended syntax and advanced modeling capabilities.

Once the model status is complete, start making predictions either one by one or in bulk:

method 1 (one by one):

SELECT predicted_column_name
FROM mindsdb.model_name
WHERE condition1='value1', condition2='value2' . . .

method 2 (bulk via JOIN):

a.input_column1, a.input_column2
FROM your_db_name.input_table AS a
JOIN model_name AS b;

Please refer to MindsDB community Slack to get support or ask questions from experts


What can I do with MindsDB?
There are many AI possibilities you can build - like predicting sales, inventory levels, assessing risks, analyzing sentiment, detecting hate speech, etc - you name it. Please check community tutorials available on and online

I'm an experienced data scientist. Can I benefit from MindsDB?
Of course. Bringing ML models to the data layer solves many fundamental machine-learning problems and significantly simplifies ML workflows.

How is my database performance affected?
MindsDB runs on a separate instance and doesn't impact your database performance

How can I fine-tune models?
Please refer to docs for an available syntax or how to import a pre-built model.

What kind of custom-built models can I import?
You can import models via MLFlow, Ray Serve, and also from NLP repositories like Hugging Face, OpenAI, etc.

What support options are available
Support is provided via community Slack There is also Enterprise Support available on request.

How can I contribute to MindsDB?
Please check community guidelines and incentives at




Operating System

Linux/Unix, Ubuntu 22.04 LTS (Jammy Jellyfish)

Delivery Methods

  • Amazon Machine Image

Pricing Information

Usage Information

Support Information

Customer Reviews