thought exercise
When I first started at Tableau, I led a project called DataHub which set a long-range vision for a project like Copilot. But without the power of GAI, we couldn't forge a path forward with our current technical abilities.
early design work
This role spanned multiple GenAI projects including an inline copilot experience.
For Copilot, my team was tasked with integrating Salesforce ChatGPT into Tableau. We started by brainstorming a blue-sky vision - thinking about all the capabilities we could offer our users with this technology that would help them see and understand their data in deeper and more meaningful ways.
communication and handoffs
Scoping this vision took many meetings, slack conversations, design doc revisions, and technical scoping.
We needed high-level sign-off quickly on this project, so we went medium-fidelity for the first mockups in order to speak to VPs in a visual way they would relate to. This mockup shows how a user generated query can be turned into a data visualization quickly and more prompts from Copilot can enable a user to dig even deeper into their data source.
other ai projects
Beyond Copilot, I led other projects integrating AI into a user's workflow to provide contextual assistance where and when it would be most helpful without taking a user out of their workflow.
The example on the right shows an early iteration of integrating AI inline when writing a calculated field while doing exploratory data analysis.
Tableau Pulse surfaces AI-driven explanations for the value of a data point with a single click. The insights are easy to share across platforms such as slack or to embed in reports and slideshows. It's optimized for mobile, but also provides a desktop version and an admin experience.
Data Stories creates automated plain-language explanations for Tableau dashboards. These descriptions are added as metadata to the data sources so that they can be further utilized by Salesforce Intelligence on the Salesforce Cloud.