Reuse knowledge for Agentic AI: Practical workshop on reducing cost & latency
Conference India 2026
November 25–26, 2026
Bengaluru, India
Joining remotely?
Watch live with ProWhat this session covers
Imagine your pager going off at 3 AM, only to find that your LLM costs are about to exceed their threshold. Generative AI applications become slow and expensive when every user request is routed directly to an LLM.
This workshop introduces the fundamentals of semantic caching, dual-memory architecture for agentic memory. We’ll explore where semantic caching fits within a modern LLM stack and cover extraction strategies for capturing what’s relevant in a session, and how to promote important facts to long-term storage.
Participants will build a working real-time AI assistant that incorporates semantic cache layer and agent memory to maintain context and remember useful user information that compounds over time.
Conference India 2026
Related talks
- Bypassing the GPU Tax: Smart Layer-7 Routing with Small Language Models (SLMs)
Sagar Utekar · CrowdStrike
Thu 26 Nov · Stage 2
- When Caches Lie: Solving Transactional Integrity and Distributed Race Conditions in Spring Boot
Shuchita Prasad · JPMorgan
Thu 26 Nov · Stage 2