Copilot Evaluation Programme
As part of the Copilot Evaluation Programme (one of UK's largest AI rollouts), I contributed from pilot readiness through to implementation, with a particular focus on the evaluation phase. Leading on user research, I gathered insights into staff experiences, adoption barriers, and measurable benefits to inform future scaling. My findings helped shape the evidence base demonstrating Copilot's impact on efficiency and productivity across NHS teams.
National Impact
This major national AI trial was featured by GOV.UK, The NHS Confederation, The Telegraph, and The Independent, and shared by leaders like Sonia Patel (NHS CTO) and Chris Parsons (NHSE).
User Research Approach
A 5-phase research journey
Use case collection
A use case form was distributed to pilot participants to identify how they use Copilot. This gathered diverse real-world applications, forming the foundation for further research.
Use case analysis & selection
Collected use cases were analysed to identify high-value examples across five personas. Selected cases were prioritised for user interviews.
User interviews
9 x 30-minute user interviews were conducted in a relaxed format to explore selected use cases. These uncovered challenges and unique applications.
Persona-based thematic analysis
Use cases within the form were grouped by personas, and Copilot identified recurring themes, offering structured insights into effectiveness.
Correlating themes and insights
Themes identified from the use case form were compared to the interview insights. This refined the understanding of impact and positive sentiment.
Research Output Headlines
Key metrics from the evaluation
Use Cases Collected
User Personas and Journeys Created
User Interviews Conducted
Unique Time Saving Use Cases Identified
High-Value Use Cases Evaluated
User Analysis Deep Dive Insights Created
Key Common Themes per Persona
Minutes of Interviews Held
User Insight Spotlight
Below showcases one of nine in-depth insights derived from user interviews, capturing the persona, role, key use cases, and powerful firsthand quotes.
