Saving Pharma Companies Billions with AI l Patrick Leung from Faro Health
Mar 21, 2025
auto_awesome
Patrick Leung, CTO of Faro Health and former leader at Google, discusses revolutionizing clinical trial design with AI. He reveals how AI transforms regulatory document generation and captures insights from past trials, showcasing the importance of collaboration between clinical experts and engineers. Leung addresses the high costs of trials, challenges with traditional metrics, and the role of structured data extraction in enhancing outcomes. He also dives into AI's capabilities and the hiring challenges faced in the field.
Faro Health leverages AI to transform clinical trial design, enabling more efficient document generation and regulatory compliance processing.
The collaboration between clinical experts and AI engineers is crucial for developing reliable systems that effectively address healthcare challenges.
Faro Health's innovative SaaS platform improves trial design by centralizing information, reducing duplication, and providing insights from historical data.
Deep dives
Challenges in Clinical Trial Design
The traditional method of designing clinical trials has become increasingly complex, leading to the phenomenon known as Eroom's Law, which indicates that the cost of bringing new drugs to market doubles approximately every nine years. Factors contributing to this trend include advancements in genetic medicine, heightened regulatory scrutiny, and the necessity for diversity in patient populations. These complications highlight the need for optimally designed clinical trials, as poor design can lead to issues such as high patient burden and protocol adherence problems. Faro Health addresses these challenges with a SaaS platform that allows for more efficient clinical trial design, ultimately reducing costs and improving the likelihood of successful outcomes.
Innovations in Clinical Testing
Faro Health utilizes a sophisticated SaaS tool that changes the way clinical trial protocols are developed, moving beyond traditional methods often reliant on Microsoft Word documents. The platform aims to provide insights early in the development process, allowing pharmaceutical companies to understand factors such as patient burden and cost complexity more effectively. By centralizing information into a structured database, the tool helps eliminate duplication and inconsistencies found in multiple documents submitted for regulatory compliance. The integration of AI into this platform is intended to further streamline the design process and enhance the quality of clinical trial documentation.
Utilizing AI for Document Generation
To facilitate the generation of required FDA documentation, Faro Health employs advanced large language models (LLMs) to ensure clinical trial protocol sections meet regulatory standards. This involves creating a dual-model system where one model generates textual content based on structured data, while the second evaluates the output against established criteria to maintain clinical quality. The evaluation system is crucial as it relies on clinical experts to refine and adapt checklists used for quality assurance, ensuring the generated documents are comprehensive and usable. This iterative process allows for continuous improvement in document generation, addressing various challenges encountered in clinical protocol writing.
Data Extraction from Existing Trials
Faro Health has also focused on utilizing AI to parse and extract structured data from existing clinical trial documents found in public repositories like clinicaltrials.gov. This involves developing algorithms to handle various formatting issues inherent to PDF files, ensuring that key assessments and schedules are effectively normalized and cataloged. By automating the extraction process, the tool can significantly reduce the time and effort typically needed for researchers to find relevant benchmark trials. Ultimately, this capability enhances the design of new trials by providing users with insights based on historical data, leading to better-informed and optimized trial structures.
Building a Capable Team
The development of Faro Health's AI-driven platform necessitates a diverse team with expertise in both clinical science and machine learning, highlighting the challenge of finding individuals with real-world experience in building LLM products. The team must include full-stack engineers, data scientists, and clinical experts who can collaboratively innovate and enhance the trial design process. Emphasizing the importance of practical experience over theoretical knowledge ensures that the team can execute complex data-driven tasks effectively and responsibly. As the field evolves, continually refining the selection criteria for new hires based on these insights will be essential for sustaining meaningful progress.
In this episode of High Agency, Patrick Leung from Faro Health explains how they're using AI to revolutionize clinical trial design by both generating regulatory documents and extracting insights from thousands of existing trials. Patrick emphasises the essential collaboration between clinical experts and AI engineers when building reliable systems in healthcare's high-stakes environment.
Chapters: 00:00 - Introduction 04:26 - Clinical trials before: Microsoft Word Documents 08:17 - Document generation using AI 12:26 - What makes clinical trials so expensive 16:26 - Parsing and processing clinical trial data 18:04 - Challenges with traditional evaluation metrics 21:28 - Importance of domain experts in the evaluation process 24:35 - Collaboration between domain experts and engineering 31:26 - Building a graph-based knowledge system 34:27 - Roles and skillsets required 38:06 - Lessons learned building LLM products 40:56 - Discussion on AI capabilities and limitations 46:07 - Is AI overhyped or underhyped
------------------------------------------------------------------------------------------------------------------------------------------------ Humanloop is the LLM evals platform for enterprises. We give you the tools that top teams use to ship and scale AI with confidence. To find out more go to humanloop.com
Get the Snipd podcast app
Unlock the knowledge in podcasts with the podcast player of the future.
AI-powered podcast player
Listen to all your favourite podcasts with AI-powered features
Discover highlights
Listen to the best highlights from the podcasts you love and dive into the full episode
Save any moment
Hear something you like? Tap your headphones to save it with AI-generated key takeaways
Share & Export
Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more
AI-powered podcast player
Listen to all your favourite podcasts with AI-powered features
Discover highlights
Listen to the best highlights from the podcasts you love and dive into the full episode