Sifflet AI Chatᴮᴱᵀᴬ

Sifflet AI Chat is your collaborative partner for data reliability. It helps you navigate your data ecosystem, understand the platform, and assess the health of your data through natural conversation.

ℹ️
This feature is currently available to Enterprise plan customers and a limited group of beta testers.
If you would like to learn more , please contact our Sales team.

Business Context & Data Insights

Use Sifflet AI Chat to get a comprehensive overview of your organization’s data health, governance, and accountability. For data insights, you can also ask Sifflet AI to generate reports and download them directly from the chat.

How to Download AI Responses

Click on the three-dot menu (⋯) in the AI message, or type /Downloads
Select Download
Press Enter
A secure link will be generated for you to retrieve the file

Example Use Cases & Prompts

What you want to know	Example Prompts
Governance Gaps	“List all assets or incidents missing owners.” “Who is assigned to the oldest open incident?”
Team Performance	“Which team or user owns the most failed datasets today?” “What is the average resolution time for incidents?”
Health Trends	“What are the most frequent issues this month?” “What’s the overall health of our data ecosystem today?” “What anomalies were detected in the last 48 hours?”
Proactive Advice	“Identify assets without monitors and suggest which ones I should create.”

Platform Guidance

If you’re unsure how to perform a task in Sifflet or need help with technical configuration, the AI acts as a live documentation assistant.

What you want to know	Example Prompts
Monitor Setup	“How do I create a Freshness or Volume monitor?” “How do I write a cron expression for a monitor at 6:00 AM UTC?”
Integrations	“How do I connect my Snowflake data source?”
Concept Learning	“Explain the difference between [Monitor Type A] and [Monitor Type B].”

Security, Access & Permission Rules

Sifflet AI Chat strictly respects your organization’s security model.

AI responses follow user roles, domain-level access, and tenant permissions.
Users cannot access information outside their authorized scope.
All responses are limited to the user’s permitted data visibility.
Data is transmitted securely and is never used to train external models.
Sifflet AI Chat uses tenant metadata only and does not process sensitive underlying data.

LLM Provider Data Usage

Sifflet AI leverages Anthropic’s API. Anthropic states that data sent through their API is not used to train their language models. You can review their commercial data policies here.