
Recsperts - Recommender Systems Experts
#17: Microsoft Recommenders and LLM-based RecSys with Miguel Fierro
Jun 15, 2023
Miguel Fierro, a Principal Data Science Manager at Microsoft with a PhD in robotics, dives deep into Microsoft's open-source recommenders repository, which boasts over 15k stars. He reveals how he transitioned from robotics to personalization, explaining the critical components of the system: examples, library, and tests. The conversation also explores the transformative impact of LLMs on recommender systems and emphasizes the ethical challenges and biases that must be addressed. Fierro concludes with insights on being a T-shaped data professional to thrive in a competitive landscape.
01:02:59
Episode guests
AI Summary
AI Chapters
Episode notes
Podcast summary created with Snipd AI
Quick takeaways
- The Microsoft Recommenders repository exemplifies successful open-source collaboration, showcasing over 30 algorithms and 900 tests to enhance personalization.
- Miguel Fierro emphasizes the importance of T-shaped professionals, combining deep expertise in recommender systems with broad knowledge of MLOps and business metrics.
Deep dives
The Landscape of Recommender Systems
Many companies claim to have recommender systems in place, yet a deeper examination often reveals otherwise. This discrepancy suggests that while the technology exists, true implementation might be lacking, which can hinder user experience. Effective recommender systems have demonstrated a remarkable return on investment (ROI) by enhancing revenue and engaging customers more effectively. A significant portion of Amazon's revenue, for instance, is attributed to its recommendations, highlighting the financial benefits that well-designed recommender systems can offer.
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.