Briya
From AI feature to core value proposition
Over the past 16 weeks, Flexxi is transforming Briya's rigid Population Builder into a single, polished conversational flow.âš
By marrying our deep experience designing premium AI products, we delivered a unified interface, consolidated scattered functionality, and laid a clear path toward Briya's longâterm goal: a full AI research assistant.
The Challenge
Briya is a secure healthcareâdata platform connecting lifeâscience researchers to deâidentified patient records. Briya's AI was already translating freeâtext queries into patient populationsâbut only on one specialist screen. Researchers still had to:
- Hop across legacy modules to tweak criteria
- Wrestle with an inconsistent UI
We unlocked Briya's AI researchâassistant vision by expanding the narrow NLP proofâofâconcept into a streamlined conversational research experience, unifying scattered tools, and beyond, into one intuitive flow.
First milestone
We rebuilt Briyaâs population builder, laying the groundwork for their vision
of a conversational research assistant.
Empty State
The refreshed interface opens with generous whitespace and refined typography that draw focus to what matters most. An âAll Patients Explorationâ scope selector keeps users grounded in the dataset theyâre examining, while a curated set of quick-start prompts eliminates blank-slate anxiety and gets researchers exploring insights within seconds.
Population Generated
Building on that clean canvas, we introduced flexible, color-coded entity chips: Condition, Observation, Medication, and more, paired with instant KPIs for patient count, age, and gender. Every query adds to a running Populations History to retrace analytical steps, and contextual âResearch Companionâ cards surface AI-generated insights exactly where they're needed, without forcing users to jump between views.
North Star Preview
In 3 months we made Briyaâs vision for an AI research assistant a tangible reality.
Conversational Assistant
Part of Briya's vision, a multi-thread chat layout lets researchers explore several study questions in parallel. Rich natural-language answers appear as expandable population cards, and users can drop in CSVs or imaging metadata to ask context-aware questions on the fly. Built-in guardrails: privacy labels, error disclaimers, latency indicators, and explicit data-scope cues, ensure transparency and trust at every turn, while inline conversation tools and a âSaved Itemsâ stack make it effortless to bookmark findings and continue the dialogue later.
Key Outcomes
- Boosted deal close rate by ~50 % with the new Population Builder demo, now showcased worldwide.
- Implemented practical, research-ready AI UX patterns that resonate with medical-data workflows.
- Delivered a hybrid interface that blends LLM-powered conversation with clear, task-oriented controls.
- Defined the roadmap for a fully conversational AI Research Assistant, aligning stakeholders on vision and milestones.
- Presented an impactful strategy that galvanized leadership support.
- Showcased Flexxi's edge in strategic design, rapid execution, and tight client collaboration.
Looking Forward
Flexxi will continue partnering with Briya to hone the AI Research Assistant vision via user and stakeholder insights, embed AI capabilities in Briyaâs strategic roadmap, and pioneer advanced interaction models that keep Briya at the forefront.
