> For the complete documentation index, see [llms.txt](https://docs.soda.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.soda.io/soda-documentation/soda-ai/overview.md). # Overview Discover how Soda AI can help you improve and automate your data quality seamlessly

Feature	What it does
Chat interface	Conversational entry point to Soda AI: performs actions in Soda, guides workflows, and answers questions, all in-product.
Contract Autopilot	Creates a first version of a data contract.
Contract Copilot	Helps edit and iterate on existing data contracts.
MCP	Connect your own agents to trusted data.
Root cause analysis (RCA) agent	Investigates failing checks from Claude Code and finds the root cause.
API	Integrate Soda with your own systems.
Soda CLI	Run data quality programmatically in your existing pipelines.

**Soda AI gets your data AI-ready**. It includes several capabilities that serve one goal: trusted data that's ready for an AI-driven workflow. With Soda AI, your team and AI agents can turn data into actionable, understandable and trustworthy information. The following capabilities make this work:

Feature	What it does
Soda MCP	Connect your own agents to trusted data.
Root cause analysis (RCA) agent	Investigates failing checks from Claude Code and finds the root cause.
Soda AI Agent	Conversational entry point to Soda AI: performs actions in Soda, guides workflows, and answers questions, all in-product.
Contract Autopilot	Creates a first version of a data contract.
Contract Copilot	Helps edit and iterate on existing data contracts.
API	Integrate Soda with your own systems.
Soda CLI	Run data quality programmatically in your existing pipelines.

## What is Soda AI? **Soda AI is the collection of AI-powered features** that make your data quality AI-ready. It isn't a single tool but the umbrella over everything in Soda that uses AI to define, maintain, and reason about data quality: the **Soda AI Agent**, **Autopilot**, **Copilot**, and the **tools** that connect your contracts and data-quality status to **external agents**. Soda AI's features fall into a few groups:

		Cover image
Soda AI Agent	The Soda AI Agent is the in-product way of interacting with Soda AI in natural language. It answers questions about Soda and your data, guides you through platform workflows, and proposes/performs actions on your behalf. As a Soda AI interface, the Agent reaches for the tools below to carry out tasks.	https://1123167021-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FA2PmHkO5cBgeRPdiPPOG%2Fuploads%2FGayhWCvMT0CwU9XbSPl3%2Fsoda_chat.gif?alt=media&token=623faa81-59a0-4984-9cd9-a2e00eeaf5b5	Soda AI Agent
Contract Autopilot	Generate fully populated contracts with recommended checks, drawn from your data's own profile. Data teams can reach broad coverage in an afternoon instead of authoring contracts by hand for weeks.	https://1123167021-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FA2PmHkO5cBgeRPdiPPOG%2Fuploads%2FbB6j0IuYb8PxlkA3Vh6W%2Fsoda-autopilot.gif?alt=media&token=304cc243-5968-4706-b83f-b954d23f46dc	Contract Autopilot
Contract Copilot	Edit and iterate on contracts in plain English (adding a freshness check, tightening a rule, handling a nested field) without hand-writing code lines. Contract Copilot will show you exactly what changed.	https://1123167021-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FA2PmHkO5cBgeRPdiPPOG%2Fuploads%2FH1WDk6863pSYzVg8LVWg%2Fsoda_copilot_short.gif?alt=media&token=bad2660c-1d07-4ec5-a50c-bf7d08830fb8	Contract Copilot
Agent-facing interfaces	Connect agents to trusted data. MCP, Soda CLI and API put your contracts and data-quality status in front of the tools your team and your agents already use. Engineers can run quality checks inside their existing pipelines, and agents act on verified data.	https://1123167021-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FA2PmHkO5cBgeRPdiPPOG%2Fuploads%2Fibl0Rsrn6CAGgsTkEjub%2FFrame%203518(4).png?alt=media&token=fce24d27-42c5-4dc6-8dc2-72c9854d7b43	Soda MCP

Across all of these, **Soda AI is designed to**: * **Reduce manual effort** in defining and maintaining data quality expectations * **Help** users **reason about data** and contracts faster * **Integrate with existing workflows**, enabling AI agents to interact with Soda Cloud in your own pipelines. * **Keep humans in control**: every proposed action is shown and approved by a person before it is applied * **Operate safely** within strict privacy, security, and governance boundaries ### Limitations * Soda AI features may not always produce complete or perfectly accurate answers. * Users remain responsible for reviewing any proposed action before it is applied. * The Soda AI Agent, Contract Autopilot, and Contract Copilot do not operate outside the context of your Soda Cloud workspace. {% hint style="success" %} **The Soda AI Agent, Contract Autopilot, and Contract Copilot are assistive, not autonomous**. These features are designed as a control surface for human-in-the-loop data quality, scoped to the same operations you could perform yourself, subject to your role and permissions. {% endhint %} *** ## Enable/Disable AI features AI features in Soda can be configured from the **Organization Settings**. {% stepper %} {% step %} Navigate to **Settings** > **Organization Settings**. {% endstep %} {% step %} Under the **Organization** tab **> Soda AI features:** Check/Uncheck the feature checkbox to enable/disable it.

{% endstep %} {% step %} Click **Save** on the top right corner. {% endstep %} {% endstepper %} *** ## Soda AI security & privacy measures Soda's features are built with **privacy and security as first-class design principles**. The following safeguards apply to all AI/ML-powered features in Soda. ### Security and privacy Soda AI features follow the same security and privacy principles as the rest of Soda: * **Your data isn't used to train models.** Neither Soda nor the model provider trains or fine-tunes models on your data, metadata, or results. * **Source data isn't sent by default.** Row-level data and PII only reach a model provider for features you opt in to. * **Soda AI acts within your permissions.** It can't do anything you couldn't do yourself with your role. * **Failed rows stay in your warehouse.** Rows stored by Diagnostics Warehouse aren't available to Soda AI. > For what each feature sends to the model provider, what is stored, and which settings admins control, see [AI security, compliance & privacy](/soda-documentation/soda-ai/ai-security-compliance-and-privacy.md). #### **Confidentiality & access control** * Soda AI's actions are subject to your **existing** **Soda roles and permissions**. None of the features can perform anything you could not perform yourself. * Requests are **not used for model training**. ### AI guardrails and ethics * Foundation models include built-in safety guardrails from the provider * Additional **system-level guardrails** restrict responses to the intended Soda use cases * Soda AI features do not generate harmful, discriminatory, or user-targeted content Because Soda AI features do not evaluate users or make decisions, traditional bias and fairness risks are minimal.
*** {% hint style="info" %} You are **not logged in to Soda** and are viewing the default public documentation. Learn more about [Licensing & documentation access](/soda-documentation/govern/security-licensing-and-compliance/documentation-access-and-licensing.md). {% endhint %} --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.soda.io/soda-documentation/soda-ai/overview.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.