Perplexity’s Computer now connects to Snowflake, letting teams ask natural‑language questions against enterprise tables to surface metrics and act on insights without waiting on SQL, dashboards or analyst queues.
Perplexity has released an App Connector that links its Computer product to Snowflake, enabling employees to query enterprise data in plain English instead of relying solely on SQL, dashboards or long requests to data teams. The integration translates natural‑language prompts into analytics results against a company’s Snowflake instance, moving analysis directly into team workflows and reducing turnaround time for routine data questions.
Once a Snowflake connection is configured, Computer can convert plain‑language questions into concrete analytics outputs. The company highlights use cases such as diagnosing the top drivers of support tickets, forecasting eight‑week demand to adjust staffing and inventory, identifying the fastest‑growing customer segments, and comparing conversion performance across marketing campaigns. Administrators can maintain shared business definitions so answers follow consistent logic across users.
The Connector builds on Perplexity’s internal tooling and semantic work. Engineers previously deployed a Slackbot that queries Snowflake and a shared semantic layer that maps key tables, metrics and common query translations into SQL; that internal system now runs up to 3,000 queries per week across product, analytics, growth, infrastructure, marketing, support and security. Perplexity cites that activity as evidence the approach can scale beyond the company’s internal teams.
Perplexity positions the Connector as a way to extend Snowflake‑hosted data into everyday workflows. Given Snowflake’s wide enterprise adoption, the company says the feature is designed to make stored data more accessible to nontechnical users. The Connector is available today to Pro, Max and Enterprise customers.
The integration is built to respect existing enterprise security and governance. It preserves Snowflake role‑based access control (RBAC), supports read‑only restrictions, and Perplexity recommends User OAuth so employees connect with their own Snowflake accounts. Administrators can immediately revoke access by removing roles or disconnecting the Snowflake link; Perplexity also states that data from the connection is not used to train models and that the integration includes encryption and privacy protections.
Admins manage the connection and semantic context inside Computer’s Settings under Connectors: after linking Snowflake they can click Generate data map to create a semantic context layer and then open the data map editor via View knowledge. Perplexity warns that generating a data map can take up to 90 minutes depending on the size and complexity of the data platform. For builders and data teams, the Connector aims to shorten analyst queues and provide contextualized, governed access to enterprise tables directly within product and team workflows.
Sources
Replies (0)
No replies in this topic yet.
