The AWS Developers Podcast AWS Hero Linda Mohamed: Juggling Cloud, Community & Agentic AI
Mar 11, 2026
Linda Mohamed, AWS Community Hero and independent cloud consultant, went from Java telecom work to building serverless and multi-agent AI systems. She discusses discovering Lambda via an Alexa skill, creating Otto the multi-agent Slack bot, and turning an AI video-analysis pipeline into a paid product. Conversation covers conference-driven development, vibe coding versus spec-driven work, and practical choices for agent frameworks and deployment.
AI Snips
Chapters
Books
Transcript
Episode notes
Conference Demo Became A Costly Cloud Lesson
- Linda submitted a multi-cloud juggling-ball demo to a major conference and ran out of time, ending up deploying with Docker on a single EC2 instead of the planned managed services.
- The demo taught her about cloud costs, missing operational steps, and the need for budget alerts after a weekend run ballooned costs until refunded by support.
Watch Setup Videos Before Spinning Up Cloud
- Do watch a couple of short setup videos before experimenting in cloud to avoid costly mistakes like forgotten queues or runaway processes.
- Linda learned this after a weekend test ran up unexpected charges because she didn't set budget alerts or delete processed queue messages.
Start With Lambda To Prototype Business Logic
- Start serverless experiments with Lambda and add Step Functions and infra-as-code later to focus on business logic first.
- Linda prefers Lambda for demos and recommends deciding deployment pipelines early to avoid rework.
