Join Philip Kiely (Lead Developer Advocate, Baseten) and Agustín Bernardo (Senior AI Engineer, Superhuman) for a deep dive into how embedding models power Superhuman’s AI-native email client.
What you’ll learn:
Which product features embedding models make possible (not just RAG)
How embedding models fit into multi-step, multi-model AI pipelines
The tradeoffs between open- and closed-source models
Selecting, evaluating, and fine-tuning embedding models for your use case
How latency, token cost, and TCO shape architecture decisions
Embeddings are the backbone of semantic search, personalization, and retrieval at scale. Hear how Superhuman chose models, balanced latency with accuracy, and optimized user experience. If you’re building an AI product in or for enterprise, you'll walk away with practical advice to apply in production.
