01
SuperLiving
Early-stage AI systems
Contributed to core AI systems during the company’s early stage, exploring how models and context come together in a product.
AI × ProductEngineer’s curiosity. Founder’s mindset.
Ideas to systems.
Systems to impact.
I’m Soham. I explore the space between technical insight and useful products—building, asking better questions, and sharing what I learn along the way.
THE FOUNDATION
MeeshoData Scientist
IIT BombayMechanical Engineering
AI → ProductThe thread connecting it all
01 / From ideas to implementation
Learning by building.
Selected chapters across AI products,
early-stage teams, and local intelligence.
02
Srota
Local intelligence
Worked on the idea of local intelligence: bringing capable, private AI closer to the people and devices using it.
Edge AI03
IIT Bombay
AI community
Part of a small team building a retrieval-augmented AI agent to make institute-specific information accessible to students.
RAG × Agents02 / Thinking out loud
Notes from the frontier.
Ideas in progress. Lessons from building.
A public record of changing my mind.
Essay
Learning in public, beyond the easy explanations
On self-teaching, following questions into uncomfortable depth, and balancing understanding with action.
Research direction
Following an AI system across the whole stack
A map of the territory—from distributed systems and reliability down to kernels, memory, and hardware.
Research note
Why local intelligence matters
What becomes possible when capable models run privately and efficiently on ordinary devices?
03 / What comes next
Questions I’m exploring.
A research agenda across intelligence,
infrastructure, and entrepreneurship.
Adaptive intelligence
Test-time training, continual learning, and systems that keep evolving after training.
Systems at scale
Distributed systems, reliability, and the journey from a promising prototype to a dependable product.
Efficient inference
CPUs, GPUs, kernels, and the architecture that brings intelligence onto everyday devices.
Building a company
How insight becomes a product, demand is discovered, and distribution turns invention into impact.
04 / Beyond the work
A work in progress.
With a bias to build.
From Mechanical Engineering at IIT Bombay to AI systems and early-stage teams, my path has been guided by curiosity. The next chapter is about turning that understanding into something people find useful.
More about meUnderstand enough to build. Let reality expose the gaps. Return with better questions.