(Overview) AstraZeneca: Agentic AI Quality-Control System for Pharma

Business Context & Design Challenge:

With numerous pharmaceutical production sites around the world, AstraZeneca was looking for a way to increase the efficiencies of those complex processes while maintaining best-in-class quality. With a previous, labor-intensive version of a quality-control system in place and performing below expectations, this project required a careful balancing of previously established workflows with fresh, innovative thinking.

I was brought aboard during a bogged down project conceptualization phase, and developed a mandate to drive research and product strategy, facilitate ideation with a cross-disciplinary team, and lead the iterative design of the final product.


Project Profile & UX Strategic Approach:

Carrying a high amount of complexity and risk in this 0-to-1 MVP project, I worked with an initially small team to implement a Hybrid UX strategy consisting of a burst of intensive research before leading a larger cross-functional team to design and deploy the initial MVP, and followed by a fully featured solution.


Competency Highlights & Results:

Quickly established cross-functional alignment to deliver an industry-first solution in less time:

  • Established shared product vision to deliver MVP in 3 months (4 weeks ahead of schedule).
  • Delivered a complete tiered solution in 6 additional months (3 months ahead of schedule).

Leveraged strategic design expertise to introduce novel capabilities that increased production efficiency while reducing costs:

  • $1M per month savings (across 6 sites) via 20% reduction in unscheduled downtime, 15% reduction of materials wastage.
  • Increased issue diagnosis speed by 80%.

Provided thought-leadership, guidance, and mentorship for expanded product team for full-scale North American deployment:

  • Worked with Product Owner and SMEs to update executive stakeholders, and develop system implementation protocol for North American adoption.
  • Facilitated on-going design reviews for 11 direct reports to seamlessly integrate AI functionality into system interface.