Technical posts on LLMs, AI agents, production ML systems and more. I write about lessons learned from building and playing with AI/ML, as well as general technology.
Google ADKFastAPIServer-Sent EventsProgressive LoadingBackground JobsReal-time Updates
How to extend Google ADK's FastAPI server with custom endpoints for progressive data enrichment, integrating async background jobs with Server-Sent Events streaming.
How to improve Lambda cold start performance by 56% through profiling, lazy imports, and strategic memory allocation - making it actually cheaper to run.
Google ADKReactTypeScriptServer-Sent Eventsfrontendstate management
How to built a production-ready React frontend for Google ADK agents with Server-Sent Events streaming, structured data rendering, and multi-agent state management.
Google ADKretry strategieserror handlingproductionreliabilityrate limit
How robust retry strategies with exponential backoff in Google ADK can handle rate limits and network failures automatically, improving production reliability.
Google ADKAWSECSDockerdeploymentinfrastructureintegration
How I deployed Google ADK agents to AWS ECS to integrate with existing systems and databases, moving away from Google Cloud defaults while maintaining full functionality.