I was engaged to help conceptualise an idea to create a natural-language data querying tool for a specialised business insights platform.

FireflyBI was pioneering a new data processing and business insights platform used by organisation with complex, fragmented data source.

My deep knowledge of UX/UI and working with complex requirements was instrumental in conceptualising a GPT-style ‘chat’ interface for non-technical users to ‘talk to data’. We named it Firefly Lumi – a riff on the idea of illuminating data in new ways.

Firefly Lumi was conceived before ChatGPT or other chat-based AI tools existed, so there were no good reference in the market to use at the time.

The primary challenge was to find an intuitive approach to querying insights across a complex data set and get simple, understandable insights back which could then be used to build reports or refine.

Leveraging my expertise in user experience and interaction design, I took a multi-faceted approach to create the concept:

Architected a series a typical request/response playbacks

Designed a chat-based UI for data interaction with simple ‘ask to start’

Designed framework for history and ‘pick up where you left off’

Designed support mechanics for collaboration, integrations, and reporting

The focus on intuitive user experience and high-fidelity concepts helped the Firefly team to:

Secure additional funding

Gain feedback from current clients

Inform their roadmap for the next year

Maintain a position of market innovation