Hello there
I am Govind, a passionate technologist and lifelong learner who loves exploring new ideas, systems, and ways of thinking. I love thinking - especially thinking about thinking!
I am
Govind KP, a passionate technologist and lifelong learner who loves exploring new ideas, systems, and ways of thinking. I love thinking - especially thinking about thinking!
I’m a technology generalist with experience across industry and academia, bringing together academic rigor, design thinking, and practical engineering. You can learn more about my professional background for detailed information about my experience, skills, and projects.
Get in touch with me with the links at the bottom!
Research
Currently, I’m based in Rennes, France, working on my doctoral studies at Inria. If everything goes swimmingly, I’ll be defending my thesis in March 2026.
My thesis, titled “Energy-Efficient Data Stream Processing”, focuses on understanding and optimizing the energy consumption of stream processing systems like Apache Kafka. The work is in collaboration with OVHcloud as part of the FrugalCloud initiative.
AI-Augmented Memory Systems
As we get closer to machines that can plausibly think, I have explored deeply what works and what doesn’t work and what the tradeoffs are.
I have been using Whisper Speech-to-Text to transcribe everything I say and hear, processing it with simple LLM agents and integrating it with various systems like Google Calendar, Todoist, Notion, Obsidian.
My fundamental philosophy is that LLMs should ideally be asynchronous and unobtrusive. They’re really good at organizing information and filling in the gaps with little context. AI is most effective when applied to aspects of life that complement your weaknesses, working quietly in the background to enhance your capabilities rather than replacing your thinking.
I believe that using LLMs in an “active” dialogue is sub-optimal. My experience is that a dialectic model tends to slow down and constrain your thinking.
Infrastructure & Automation
One of the most fun (and sometimes frustrating!) tendencies that has changed my trajectory is my stubborn desire to find out how to do things by myself with the resources I have. The hard way is the way that teaches you the most. I’ve self-hosted my own cloud for the past 5 years, with my own storage, AI models and storage on my own infrastructure, managed with Kubernetes.
By making many mistakes and wasting many hours, this has taught me how systems work more than working with massive data center machines.
To me, there’s something deeply satisfying about having complete control over your digital environment and reducing dependencies on external services. We’re in the golden age of self-hosting, with extremely good open-source and self-hosted alternatives to almost any service you can imagine!