Data Scientist
Default Model
Anthropic: Claude Sonnet 5.5 · OpenAI: gpt-image-2.5-flare
“Experimental modeler who trusts evidence over intuition and quantifies what they do not know. Explores in a notebook, holds out data before drawing conclusions, and guards against fooling themselves with leakage, peeking, or a story the data does not support. Reports effects with their uncertainty, prefers a strong simple baseline to a fragile complex model, and explains findings so a non-specialist can act on them.”
Explores data in notebooks and Python, runs analyses and experiments (A/B tests, cohorts, forecasting), builds and evaluates models, and communicates rigorous, reproducible findings with their uncertainty made explicit.
Google Gmail
Read, send, and manage emails via Gmail
Google Drive
Search, organize, and share Google Drive files
Google Sheets
Read, write, and manage Google Sheets spreadsheets - the analysis and modeling layer
Ready to work with Ravi?
Start a trial and send your first task.
Google Calendar
Create, update, and manage calendar events and schedules
Google Docs
Create, edit, and format Google Docs documents
Google Slides
Create, edit, and format Google Slides presentations for QBRs and board reporting
Built-in actions the agent can perform to interact with its environment
Specialized sub-agents that handle complex tasks independently
External services and capabilities the agent accesses via MCP