Speak at The Fifth Elephant 2026 Annual Conference

Describe your session in 2 paragraphs

I have worked with my Al agent for two years. I gained two years of experience. My agent gained zero. That same agent worked with ten people on my team. Combined how can we give our agent 10-20 years of experience. How do we design this?

In this session, we will show you how to build the intelligence layer that makes this possible. Memory that evolves with every interaction, learns from every failure, and compounds experience over time. The goal is simple: self-evolving agents that get better the longer they run.

Mention 1-2 takeaways from your session

How to build a memory system that evolves with every interaction - compounding experience across sessions, users, and tasks instead of resetting to zero.

How to use that memory as the foundation for self-evolving agents that get smarter over time, without retraining, without fine-tuning, just by doing the work.

Which audience is your session going to be beneficial for?

Engineers and architects building or evaluating autonomous agent systems.
Practitioners working with Al system who are hitting limits on complex multisession tasks.
Technical leads responsible for enterprise Al deployments who need a principled framework for memory architecture decisions.

Add your bio

Sparsh Jain is the leading the Tech at Central Agent, building Principal Engineering Brain. Ex- Al Lead at Key Al, where he build agentic security infra with CISOs. Previously AVP of Quantitative Al at Wells Fargo. Published Al research at NTU Singapore. Alumnus of BITS Pilani. He leads Al Tinkerers Hyderabad and has spoken at Latent Space, Salesforce, and multiple universities on agentic systems and Al infrastructure.

Previous Talks:

1. Latent Space: Memory Infra (https://www.youtube.com/watch?v=XLUpkWdZLIQ)
2. Salesforce: (https://www.linkedin.com/feed/update/urn:li:activity:7466534000447361024/)

Draft Slides: (https://drive.google.com/file/d/1kFg6dxRAE-2q44IUdsVkbCnyqDxRAODD/view?usp=drive_link)

Talk Pitch: https://youtu.be/VWtdllDUwiY