Insight & Foresight

AI-enhanced interface for interpreting complex data relationships in research

 
BioGraph - AI/ML hero asset

Overview

Researchers working under grant and publication pressure faced fragmented planning workflows across multiple tools and datasets. Complex protein associations and large product catalogues increased planning time and the risk of missed research opportunities.

Risks

  • Reducing planning time without disrupting existing high-converting journeys

  • For the business, limited insight into emerging research trends constrained proactive portfolio investment.

  • Introducing new digital capability also carried risk to revenue-critical search journeys.

Priorities

  • Reducing planning time without disrupting existing high-converting journeys

  • Introducing AI as an augmentation, not a replacement

  • Prioritising speed, stability, and adoption over feature depth

  • De-risking impact through phased exploration and clear success metrics

Direction

  • Reducing planning time without disrupting existing high-converting journeys

  • Limited visibility into emerging research relationships constrained proactive portfolio investment

  • Introducing new digital capability carried risk to revenue-critical search flows

  • Designed using the BiOTA 3.0 design system to ensure brand and interaction consistency

  • Full-screen, mobile-optimised layout to simplify navigation within complex relational data

  • Visual complexity reduced from early proof of concept to avoid unnecessary cognitive noise

 

Outcomes

Researchers were able to explore pathways and product associations previously hidden across disparate sources. Planning shifted from manual, multi-tool workflows to a single exploratory experience, improving confidence and breadth of consideration.

Impact

  • Demonstrated feasibility of AI-driven planning within existing channels

  • Established a foundation for scalable digital services alongside physical products

  • Improved collaboration across Product, UX and Data Science

  • ~75% reduction in average planning time (≈3 weeks → 3–5 days)

Learnings

  • Predictive planning must enhance trusted search behaviours

  • Novel interaction models require careful onboarding and terminology clarity

  • Phased rollout is critical to adoption and risk management

 
 

Phase 2 - Planning