Karthikeyan Rajendran
Lead Engineer at Amazon
I've spent 15+ years building and shipping software that serves millions of users. From consumer products like Kindle Fire HD to infrastructure platforms like AWS Greengrass, I've learned that great engineering isn't just about writing code—it's about building systems that are reliable, scalable, and maintainable.
Today, I'm at the intersection of traditional software engineering and generative AI, building agentic systems that solve real-world problems with LLMs.
What I Do
As a software engineer, I focus on distributed systems, quality engineering, and now—agentic AI workflows.
Today, I'm building agentic systems and LLM workflows that automate complex software engineering tasks. I design multi-agent architectures that handle everything from code generation to testing and deployment, working with prompt engineering and RAG (Retrieval-Augmented Generation) systems to solve real-world problems. This isn't just experimentation—it's applying AI to the messy, practical challenges that engineering teams face every day.
This work builds on 15+ years of distributed systems engineering at AWS and Amazon. I've built and scaled backend services that serve millions of users, from consumer products like Kindle Fire HD to infrastructure platforms like AWS Greengrass. The rigor I learned building reliable distributed systems—thinking deeply about failure modes, error handling, and observability—directly informs how I approach AI systems. An LLM-powered agent needs the same careful architecture as any distributed service.
Throughout my career, quality engineering and test automation have been central to how I build software. I design frameworks and tooling that catch issues before they reach production, and I'm particularly interested in how AI agents can enhance traditional testing approaches. Intelligent test generation, automated debugging workflows, code review assistance—these are areas where AI can genuinely make engineering teams more effective, not just faster.
I've also spent years building developer tooling and infrastructure that makes teams more productive. From internal tools to AI-powered workflows for code reviews, documentation generation, and system design validation, I focus on creating leverage—building systems that amplify what great engineers can accomplish.
I hold a Master's in Computer Applications from Anna University Chennai, where I developed the foundation for building robust, well-tested systems—principles that now guide how I architect reliable AI agents.
Beyond the Code
From Soorasamharam (2023) — served as stunt director, choreographing all silambam sequences.
Teaching Silambam
I practice and teach this traditional Tamil martial art, which requires the same precision and systematic practice that makes great software. The discipline of perfecting a form through thousands of repetitions mirrors both the iterative process of building reliable systems and the careful prompt refinement needed for effective AI agents.
Mentoring Future Engineers
Through hope3.org, I mentor college students and engineers from rural India, guiding them as they build their careers in technology. With AI rapidly changing our field, I help them navigate both traditional software engineering fundamentals and emerging AI capabilities—ensuring they're prepared for the future of our industry.
Overlooking Snow Lake near Snoqualmie, WA.
Finding Clarity Outdoors
As an avid hiker exploring the Pacific Northwest, I've learned that the best architectural insights often come when you step away from the keyboard. Whether designing a multi-agent system or debugging a distributed service, complex problems often reveal their solutions when you can see the whole landscape.
What Drives Me
The most exciting problems in software engineering today sit at the intersection of traditional systems thinking and generative AI. I believe AI agents aren't replacing software engineers—they're powerful tools that amplify what great engineers can build. My work focuses on making these tools reliable, practical, and accessible.
Whether I'm building a new agentic workflow, designing test infrastructure, or reviewing someone's LLM integration, I bring the same conviction: software engineering is a craft that demands both technical depth and disciplined practice—and that's as true for AI systems as it is for any other code we write.
Let's Connect
If you're building with AI agents, care about reliable systems at scale, or are navigating your own path in software engineering, I'd love to connect.
All opinions are my own.