Before we start, install MindsDB locally or sign up for MindsDB Cloud.

Get started with MindsDB in a few simple steps:

1

Connect a data source

Explore all available data sources here.

2

Configure an AI engine

Explore all available AI engines here.

3

Create and deploy an AI model

MindsDB abstracts AI Models as AI Tables. This step uses the configured AI engine.

4

Query for predictions

Join the data table with the AI table to get predictions.

5

Automate customized workflows

Use Jobs or Triggers to automate workflows.

Step 1. Connect a data source

Use the CREATE DATABASE statement to connect a data source to MindsDB.

CREATE DATABASE mysql_demo_db
WITH ENGINE = 'mysql',
PARAMETERS = {
    "user": "user",
    "password": "MindsDBUser123!",
    "host": "db-demo-data.cwoyhfn6bzs0.us-east-1.rds.amazonaws.com",
    "port": "3306",
    "database": "public"
};

This is the input data used in the following steps:

SELECT *
FROM mysql_demo_db.questions
LIMIT 3;

Here is the output:

+------------------+--------------------------------------------------------+-------------+
| article_title    | question                                               | true_answer |
+------------------+--------------------------------------------------------+-------------+
| Alessandro_Volta | Was Volta an Italian physicist?                        | yes         |
| Alessandro_Volta | Is Volta buried in the city of Pittsburgh?             | no          |
| Alessandro_Volta | Did Volta have a passion for the study of electricity? | yes         |
+------------------+--------------------------------------------------------+-------------+

Step 2. Configure an AI engine

Use the CREATE ML_ENGINE command to configure an AI engine. Here we use the OpenAI engine.

CREATE ML_ENGINE openai_engine
FROM openai
USING
	openai_api_key = 'your-openai-api-key';

Step 3. Create and deploy an AI model

Use the CREATE MODEL statement to create, train, and deploy an AI model based on the AI engine created in step 2.

CREATE MODEL question_answering_model
PREDICT answer
USING
    engine = 'openai_engine',
    prompt_template = 'answer the question of text:{{question}} about text:{{article_title}}';

This model expects {{question}} and {{article_title}} as input, and generates answer as output.

Step 4. Query for predictions

Query for predictions by joining the AI Table (from step 3) with the data table (from step 1).

SELECT input.article_title, input.question, output.answer
FROM mysql_demo_db.questions AS input
JOIN question_answering_model AS output
LIMIT 3;

Here is the output data:

+------------------+--------------------------------------------------------+--------------------------------------------------------+
| article_title    | question                                               | answer                                                 |
+------------------+--------------------------------------------------------+--------------------------------------------------------+
| Alessandro_Volta | Was Volta an Italian physicist?                        | Yes, Volta was an Italian physicist.                   |
| Alessandro_Volta | Is Volta buried in the city of Pittsburgh?             | No, Volta is not buried in the city of Pittsburgh.     |
| Alessandro_Volta | Did Volta have a passion for the study of electricity? | Yes, Volta had a passion for the study of electricity. |
+------------------+--------------------------------------------------------+--------------------------------------------------------+

Step 5. Automate customized workflows

With MindsDB, you can create custom automation workflows. Let’s set up a workflow that uses Jobs and (re)creates a table with predicted answers to all questions.

CREATE JOB answer_questions (

CREATE OR REPLACE TABLE data_source.questions_answers ( SELECT input.article_title, input.question, output.answer FROM mysql_demo_db.questions AS input JOIN question_answering_model AS output ) ) EVERY 1 day;

The data_source connection should be made using the CREATE DATABASE statement to a data source with a user that has the write access.

This job creates the questions_answers table inside the connected data source. This table is filled with questions from the input data table and answers generated by the AI table. Considering that new questions are added daily to the input data table, this job executes once a day.

Alternatively, you could use the LAST keyword to fetch only the newly added questions to the input data table. That would enable you to insert new question-answer pairs into the questions_answers table instead of recreating it. But to do that, the input data table must provide either a date/time or integer/float column that would be used in the condition like datetime > LAST. Learn more about the LAST keyword here.