Your lakehouse.
Now speaking C#.

The open-source .NET client for Databricks Genie.
Ask from C#, inspect returned SQL, and automate reports with Question Packs.

Start building Star on GitHub

Apache 2.0 Library: .NET 8 + 10 CLI: .NET 10

What is the total revenue by region?

The answer

FROM GENIE

Germany generated the highest revenue among the listed regions.

Revenue by region from the sanitized live transcript
region	total_revenue_eur
Germany	4500000.00
France	3350000.50
Spain	1780000.25

Generated SQL

SELECT `region`, SUM(`revenue_eur`) AS total_revenue_eur
FROM `sales`.`public`.`regional_revenue`
WHERE `region` IS NOT NULL
  AND `revenue_eur` IS NOT NULL
GROUP BY `region`

Check the query before using the answer.

Adapted from a real CLI run with synthetic data.

Inspect the original terminal output

Genie is analyzing your question…
Waiting for a SQL warehouse…
Preparing answer…
The total revenue by region is as follows:
- **Germany:** €4,500,000.00
- **France:** €3,350,000.50
- **Spain:** €1,780,000.25
Germany generated the highest revenue among the listed regions.

╭────────────────────────┬─────────────────────────────────────────────────────╮
│ region                 │ total_revenue_eur                                   │
├────────────────────────┼─────────────────────────────────────────────────────┤
│ Germany                │ 4500000.00                                          │
│ France                 │ 3350000.50                                          │
│ Spain                  │ 1780000.25                                          │
╰────────────────────────┴─────────────────────────────────────────────────────╯
╭─Generated SQL─────────────────────────────────────────────╮
│ SELECT `region`, SUM(`revenue_eur`) AS total_revenue_eur  │
│ FROM `sales`.`public`.`regional_revenue`                  │
│ WHERE `region` IS NOT NULL AND `revenue_eur` IS NOT NULL  │
│ GROUP BY `region`                                         │
╰───────────────────────────────────────────────────────────╯

View the capture source

Use Genie beyond
the chat window.

Calling Genie means managing conversations, polling, and result formats. LakeSpeak handles that in a typed .NET client, with a CLI for running questions and reports.

A typed .NET client

Dependency injection, async calls, cancellation, and typed failures.

Returned SQL and rows

Read the SQL and rows Genie returns, with decimal values preserved as strings.

Versioned Question Packs

Store recurring questions in YAML and generate Markdown reports from each run.

Ask Genie
from C#.

Register LakeSpeak and resolve IGenieClient. The client starts the conversation, waits for Genie, and returns a typed response.

Keep follow-up questions in the same conversation.
Read the SQL, rows, and message identifiers.
Cancel work with a standard CancellationToken.

Explore the runnable .NET example

Program.cs

using LakeSpeak.Genie;
using Microsoft.Extensions.DependencyInjection;

var services = new ServiceCollection();
services.AddLakeSpeak(
    options => options.Profile = "company");

using var provider = services.BuildServiceProvider();
var genie = provider.GetRequiredService<IGenieClient>();

var response = await genie.AskAsync(
    agentId: args[0],
    question: "Which regions grew fastest?",
    cancellationToken: CancellationToken.None);

Console.WriteLine(response.Text);
Console.WriteLine(response.Query?.Sql);

LakeSpeak.GenieScroll code →.NET 8 / 10

daily-platform-brief.yamlExcerpt from the repository example

questions:
  - id: expensive-queries
    title: Most expensive queries
    ask: Which queries consumed the most compute yesterday?

output:
  format: markdown
  path: reports/daily-platform-brief.md

Read the complete Question Pack

reports/daily-platform-brief.mdMarkdown output

Version questions.
Rerun reports.

Put recurring questions in a YAML Question Pack. Validate it, run each question against Genie, and write a Markdown report your team can read.

Review changes to the questions in Git, then rerun the pack as your data changes.

01 Validate02 Run03 Report

See the scheduled GitHub Actions example

Install LakeSpeak.
Connect to Genie.

Bring a Databricks workspace and an existing Genie Agent shared with you. LakeSpeak uses your identity; it doesn’t create Agents or expand your permissions.

Add the .NET library

Add the typed client to your .NET 8 or 10 project.

NuGet / library

dotnet add package LakeSpeak.Genie

Configure the .NET client

Install the CLI

Install the CLI with .NET 10, then connect below.

NuGet / global tool

dotnet tool install --global LakeSpeak.Cli

Or download a self-contained binary

Sign in to your workspace
Use the Databricks CLI for a browser-based login.
Authenticate
databricks auth login --profile company
Find your Genie Agent
List the Agents your identity can access. Copy an ID.
Discover
lakespeak agents list --profile company
Ask a question with SQL output
Replace <agent-id> with the ID from your list.
First question
lakespeak ask --profile company --agent "<agent-id>" --show-sql "What is the total revenue by region?"