2024
Dorch Copilot
dorch.co
Background
Dorch was envisioned as a productivity layer for data teams using Tableau. While Tableau offers rich visualization capabilities, analysts often spend time crafting formulas, cleaning data, and navigating dashboards manually. We built Dorch as an AI Copilot to streamline data analysis and accelerate insights inside Tableau.
Core problem
Even with Tableau, analysts struggle with repetitive tasks like writing calculations, explaining metrics, or troubleshooting dashboards. There was a need for a contextual assistant that could guide users, answer questions, and reduce manual work—without switching tools.
Steps Taken
We partnered with data analysts and BI teams to understand their daily workflows. The focus was on making the AI feel embedded within Tableau, responding to queries naturally, and augmenting—not replacing—the analyst’s flow.
Key Actions Taken
In-Dashboard AI Assistant: Built a contextual chat interface within Tableau to assist users with formulas, field definitions, and metric breakdowns.
Natural Language Queries: Enabled users to ask questions in plain English and get data-backed answers instantly.
Smart Formula Suggestions:Trained the AI to suggest Tableau formulas based on user goals—reducing dependency on documentation.
Usage Context Awareness: Made the Copilot context-aware so it understands the current worksheet and responds accordingly.
Key Features
In-app AI chat assistant for Tableau
Natural language querying for dashboards
Auto-suggested formulas and field logic
Contextual awareness across worksheets
Video of How the Copilot works in Tableau
Results and Impact
Dorch helped data teams cut analysis time, especially for repetitive queries and formula creation. It enhanced Tableau adoption within teams by making dashboards more interactive and accessible—even to non-technical users.