page_type	languages	products	urlFragment	name	description
sample	azdeveloper powershell bicep python	azure ai-services azure-openai azure-speech azure-sql-database	nlp-sql-in-a-box	NLP to SQL Chatbot in-a-box (AI-in-a-Box) with enables Azure SQL databases, Azure OpenAI, Semantic Kernel, and Azure AI Speech Service	Users to interact with SQL databases using natural language and speech, leveraging Azure OpenAI, Semantic Kernel, and Azure AI Speech Service to translate spoken queries into SQL statements, execute them, and deliver results audibly, ensuring an intuitive and user-friendly experience.

NLP-SQL-in-a-Box

This solution is part of the AI-in-a-Box framework developed by the team of Microsoft Customer Engineers and Architects to accelerate the deployment of AI and ML solutions. Our goal is to simplify the adoption of AI technologies by providing ready-to-use accelerators that ensure quality, efficiency, and rapid deployment.

User Story

Build a cutting-edge speech-enabled SQL query system using Azure Open AI, Semantic Kernel, and Azure AI Speech Service

We will use the power of Azure Open AI and Semantic Kernel to translate your natural language queries into SQL statements that can be executed against an SQL Server database. This will allow you to interact with your data in a more intuitive and user-friendly way. No more struggling with complex SQL syntax – just speak your query and let the system do the rest!

And with Azure Speech Services, we will convert your speech into text and synthesize the results as speech. This means that you can hear the results of your query spoken back to you, making it easier to understand and digest the information.

What's in the Box

• Python application that leverages Semantic Kernel and Speech Services to build a chatbot that can:
  • Understand natural language database queries from speech
  • Translate them into SQL
  • Execute the SQL against an SQL Server database
  • Return the results as speech
• Deployment templates of all resources needed, which includes:
  • OpenAI Service and Deployment
  • Speech Services
  • SQL Server
• Resources are deployed and used with security best practices in mind
  • Speech and OpenAI services do not allow api keys access
  • SQL Server requires Active Directory authentication
  • Required RBAC roles are assigned to the user deploying the solution
  • Application connects to all services using azure credential

This solution was adapted from the Revolutionizing SQL Queries with Azure Open AI and Semantic Kernel blog post.

Thinking Outside the Box

This solution can be adapted for many other use cases. Here are some ideas:

• Update the nlp_to_sql plugin to support more complex queries (including updates, deletes, etc.)
• Add more plugins to the semantic kernel to support additional use cases
• Add other options to interact with the kernel (e.g., a web interface, a mobile app, etc.)

Deploy the Solution

Deploy Pre-requisites

1. An Azure subscription
2. Install Azure CLI
3. Install Bicep
4. Install Azure Developer CLI

UI Deploy

Required Input Parameters

The parameters below are required in order to deploy the infrastructure.

• Subscription
• Region
• Environment Name
• Principal Id
  • You can find this by running the following command:

    az ad signed-in-user show --query id -o tsv

• Administrator Login

Optional Input Parameters

• IP Address
  • If you want to allow your IP address to access the SQL Server, you can provide it here.

Output Parameters

After the deployment is complete, you can find the output parameters by clicking on the Outputs tab. You need to create an .env file in the root of the project and fill it with the output parameters. The .env file should look like this:

AZURE_LOCATION="<azure_location>"
AZURE_OPENAI_CHAT_DEPLOYMENT_NAME="<azure_openai_chat_deployment_name>"
AZURE_OPENAI_ENDPOINT="<azure_openai_endpoint>"
SPEECH_SERVICE_ID="<speech_service_id>"
SQL_SERVER_NAME = "<sql_server_name>"
SQL_DATABASE_NAME = "<sql_database_name>"

Note: whenever the biceps files are changed, the azuredeploy.json file must be updated. To do this, run the following command:

az bicep build --file infra/main.bicep --outfile infra/azuredeploy.json

Azd deploy

1. Clone this repository locally

   git clone https://github.com/Azure-Samples/nlp-sql-in-a-box/

2. Deploy resources

   az login

   azd auth login

   azd up

You will be prompted for:

• environment name
• azure subscription
• azure region (we suggest using eastus2)
• database administrator login

When you deploy using this method, the .env file will be created automatically with the output parameters.

Clean up

To remove all resources created by this solution, run:

azd down

Run the Solution

Run Pre-requisites

1. Install Python 3.10
2. Install ODBC Driver for SQL Server
3. Make sure you can access the resources deployed from your local machine.
   • By default, all resources were created with no public access.
   • You can allow your own IP address to access the resources by:
     • Find out your what's your IPv4 address
     • azd env set IP_ADDRESS <ip_address>
     • azd up
4. Install requirements

pip install -r src/requirements.txt

Run Locally

python -m src.app

The first time you run the application, it will create and populate the database with fake data. This process may take a few minutes.

Logging

The application will output logs to the app.log file, so you can use it to better understand what's happening. If you need more information, you can change the log level to DEBUG app.py file:

logging.basicConfig(
    filename="app.log",
    format="[%(asctime)s - %(name)s:%(lineno)d - %(levelname)s] %(message)s",
    datefmt="%Y-%m-%d %H:%M:%S",
    level=logging.DEBUG,
)

Example Usage

Below you can see an example of the solution in action:

$ python -m src.app
Listening:
User > How many locations are there?
tool plugins-nlp_to_sql needs to be called with parameters {}
tool plugins-nlp_to_sql called and returned There are `1` tool call arguments required and only `0` received. The required arguments are: ['input']. Please provide the required arguments and try again.
tool plugins-nlp_to_sql needs to be called with parameters {"input":"How many locations are there?"}
tool plugins-nlp_to_sql called and returned ```sql
SELECT COUNT(DISTINCT Location) AS NumberOfLocations FROM ExplorationProduction;
```
tool plugins-query needs to be called with parameters {"query":"SELECT COUNT(DISTINCT Location) AS NumberOfLocations FROM ExplorationProduction;"}
tool plugins-query called and returned (1000,)
Assistant > There are 1000 distinct locations.
Listening:
User > Yes.
Listening:
User > Can you list me the top five locations by production volume?
tool plugins-nlp_to_sql needs to be called with parameters {"input":"biggest five locations by production volume"}
tool plugins-nlp_to_sql called and returned ```sql
SELECT TOP 5 Location, SUM(ProductionVolume) AS TotalProductionVolume
FROM ExplorationProduction
GROUP BY Location
ORDER BY TotalProductionVolume DESC;
```
tool plugins-query needs to be called with parameters {"query":"SELECT TOP 5 Location, SUM(ProductionVolume) AS TotalProductionVolume FROM ExplorationProduction GROUP BY Location ORDER BY TotalProductionVolume DESC;"}
tool plugins-query called and returned ('West Travishaven, Vietnam', Decimal('999300.73')),('Baileyville, Israel', Decimal('998248.91')),('Williamsborough, Wallis and Futuna', Decimal('997729.20')),('Lake Gabrielshire, Panama', Decimal('996433.80')),('Davidstad, Saint Kitts and Nevis', Decimal('994778.98'))
Assistant > Here are the five locations with the highest production volumes:

1. **West Travishaven, Vietnam**: 999,300.73
2. **Baileyville, Israel**: 998,248.91
3. **Williamsborough, Wallis and Futuna**: 997,729.20
4. **Lake Gabrielshire, Panama**: 996,433.80
5. **Davidstad, Saint Kitts and Nevis**: 994,778.98
Listening:
User > No.

Customize the Solution

Add More Plugins

You can add more plugins by:

1. Creating a new Python file in the src/plugins directory
2. Implementing your plugin as a class (more details in Plugins)

Reusing the Kernel

If you want to reuse this logic in another project, it is really easy, You just need to reuse the src/kernel package in your project, passing the required parameters.

How to Contribute

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.opensource.microsoft.com

When you submit a pull request, a CLA bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., status check, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact [email protected] with any additional questions or comments.

Key Contacts & Contributors

Highlight the main contacts for the project and acknowledge contributors. You can adapt the structure from AI-in-a-Box:

Contact	GitHub ID	Email
Franklin Guimaraes	@franklinlindemberg	[email protected]

License

This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow Microsoft's Trademark & Brand Guidelines. Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party's policies.

FAQ

1.  Server is not found or not accessible. Check if instance name is correct and if SQL Server is configured to allow remote connections. For more information see SQL Server Books Online. (11001)')
   

   This error is due to the fact that the SQL Server is not accessible from the machine where the application is running. Check Run Pre-requisites for more details.

---

This project is part of the AI-in-a-Box series, aimed at providing the technical community with tools and accelerators to implement AI/ML solutions efficiently and effectively.