Training Data Kumo’s Hema Raghavan: Turning Graph AI into ROI
22 snips
Jan 21, 2025 Hema Raghavan, co-founder and head of engineering at Kumo AI, previously led AI initiatives at LinkedIn, including the iconic 'People You May Know' feature. In this conversation, she unveils how graph neural networks (GNNs) revolutionize automated machine learning by transitioning from CPU to GPU. Hema explains how GNNs enhance recommendation systems, simplifying feature engineering and adapting to user preferences. She also discusses the transformation of relational data into graph formats and the critical role of explainable AI in advancing technology.
AI Snips
Chapters
Transcript
Episode notes
GPU Revolution in NLP
- GPUs revolutionized NLP by enabling neural networks to learn linguistic features automatically.
- This eliminated the need for manual feature engineering, like specifying parts of speech.
GNNs for Recommendations
- Graph neural networks (GNNs) excel at learning from interconnected data, like user behavior and content metadata.
- GNNs consider various factors, such as user preferences and content engagement, for recommendations.
Production Challenges with Feature Engineering
- Maintaining complex feature engineering pipelines in production can be challenging.
- A single change in logging can break the entire system, making debugging difficult.