Uploaded July 2026 | Updated November 2026, 2 days ago
Anoop Deoras leads applied science for AWS's Agentic AI business unit but his path started in speech compression and large-vocabulary speech recognition long before "agents" was a buzzword. In this conversation, he traces that journey and unpacks how AWS is building autonomous agents that go beyond simple task execution.
Topics covered:
How Amazon's internal Java migration project (30,000+ applications) became the seed for AWS Transform
Inside Kiro, AWS's coding agent, from bug fixes to full feature development
Why "context engineering" matters more than explicit rule-teaching for keeping agents aligned with team conventions
The shift from writing code to reviewing it, and how multi-agent critique setups are helping
Memory as a frontier problem: what agents should remember, forget, and consolidate over time
Why proactive agents, not just reactive ones, are the next frontier
Lessons from Netflix and Microsoft on personalization and sequence modeling
Whether you're building agentic systems or just trying to understand where autonomous AI is heading, this episode offers a grounded, technical view from someone shipping it at scale.
Anoop is joined by @Scale Podcast host, Francois Richard — Engineering Director responsible for the Reliability Infra at Meta.
Want to join sessions like this in person?
Register for our in-person and virtual events on our website: atscaleconference.com
Anoop Deoras leads applied science for AWS's Agentic AI business unit but his path started in speech compression and large-vocabulary speech recognition long before "agents" was a buzzword. In this conversation, he traces that journey and unpacks how AWS is building autonomous agents that go beyond simple task execution.
Topics covered:
How Amazon's internal Java migration project (30,000+ applications) became the seed for AWS Transform
Inside Kiro, AWS's coding agent, from bug fixes to full feature development
Why "context engineering" matters more than explicit rule-teaching for keeping agents aligned with team conventions
The shift from writing code to reviewing it, and how multi-agent critique setups are helping
Memory as a frontier problem: what agents should remember, forget, and consolidate over time
Why proactive agents, not just reactive ones, are the next frontier
Lessons from Netflix and Microsoft on personalization and sequence modeling
Whether you're building agentic systems or just trying to understand where autonomous AI is heading, this episode offers a grounded, technical view from someone shipping it at scale.
Anoop is joined by @Scale Podcast host, Francois Richard — Engineering Director responsible for the Reliability Infra at Meta.
Want to join sessions like this in person?
Register for our in-person and virtual events on our website: atscaleconference.com










