The Problem
Medical Affairs teams at pharmaceutical companies face a critical strategy-to-execution gap. Home office leaders lack real-time patient data and the flexibility to customize KOL scoring for therapeutic area-specific needs, trapping them in rigid, slow-moving consulting cycles that cost $500K+ per year. Meanwhile, field teams miss real-time clinical signals needed to reach "rising stars" and relevant KOLs before the competition defines the narrative.
As the Senior Product Manager, I led the development of Aperture 2.0 to bridge this gap with an AI-first platform for continuous discovery, natural language customization, and seamless CRM delivery.
Goals
- 🤖 AI-Driven KOL Discovery — Pinpoint "rising stars" across global research, trials, and congresses
- 💬 Natural Language Customization — Instantly customize KOL scoring and tiering in plain language
- 🌐 360° Field Intelligence — Surface clinical signals siloed legacy sources miss
- 🚀 Seamless Field Execution — Push governed insights directly into the MSL's CRM workflow
Development Timeline
- Identified the strategy-to-execution gap through customer research and competitive analysis
- Designed the AI discovery engine monitoring patient insights, trials, and congress data
- Built natural language customization replacing consulting workflows
- Developed the Analysis Canvas with contextual profiles and tiered segmentation
- Launched CRM integration pushing prioritized lists into field workflows
Challenges & Solutions
| Challenge | Solution |
|---|---|
| KOL lists couldn't be defended — black-box AI logic and months-old data | Every insight verifiable and grounded in the Healthcare Map with full scoring transparency |
| Custom scoring took 6 weeks and $150K+ per consulting project | Natural language customization eliminates the consulting dependency entirely |
| Fragmented data never converted into territory-specific field tasks | Dynamic, prioritized lists push directly into CRM workflows |
The Product
Rising star identification, deep contextual HCP profiles, an interactive Analysis Canvas with tiered segmentation, and a Discover feed of AI-curated congress, trial, and publication intelligence.
Use Cases
| Category | Applications |
|---|---|
| Field Strategy & Planning | Rising star identification, KOL profiling and segmentation, comprehensive HCP profiles, engagement insights |
| Evidence Generation | Rapid treatment pattern analysis, patient journey mapping with deeper patient and HCO insights |
| Scientific Exchange | Automated pre-call planning with AI briefs, scientific narrative development, targeted engagement planning |
| Compliance & Strategy | Custom KOL scoring and segmentation, seamless CRM delivery |
AI Evaluations
- 📋 Golden Datasets — Reference sets curated with Medical Affairs SMEs as objective benchmarks
- ⚖️ LLM-as-Judge Scoring — Automated grading of AI briefs, calibrated by human expert review
- 🔎 Groundedness Checks — Every claim traced back to the Healthcare Map, publications, trials, or congresses
- 🔁 Regression Evals — The suite ran on every prompt, model, and retrieval change
Failures & Lessons Learned
- ⚖️ Balancing AI Transparency with Simplicity — Early profiles were too opaque to defend internally; we invested in explainability without overwhelming users.
- 🔗 Adoption Requires More Than Features — Field teams needed far more onboarding and change management than anticipated.
- ⏳ Scope Creep in Natural Language — We ruthlessly prioritized launch use cases and deferred advanced scenarios.