Build
Production software today; thoughtful AI systems next.
Bengaluru · India
I build systems, follow questions, and keep crossing disciplines.
I’m Vipindev Shambu. Software engineering is what I do; curiosity across engineering, robotics, space, AI, and how people use technology is closer to who I am.
Building · learning · exploring
Learn
Mechanical engineering, robotics, space technology, and AI.
Understand
People, cognition, decisions, and how we live with technology.
One curious practice
Work is part of the pictureThe whole picture
“A job title explains the work. It does not explain the person following the questions.”
BuildingLearningUnderstanding
Two clear paths
What I do and who I am deserve different rooms.
One side is the professional record. The other is the curiosity, learning, and perspective behind it.
Production software and a transition toward AI systems.
Seven years across retail and education technology, with work in product interfaces, analytics, automation, and application architecture.
Enter the professional sideA curious generalist shaped by more than one field.
Mechanical engineering, robotics, space technology, software, and AI are chapters in the same habit: learning how complex systems work and how people make sense of them.
Meet me beyond the résuméIdeas
The questions I am following
Notes are where the professional and personal sides meet: software, AI, people, and the process of learning in public.
From frontend systems to intelligent agents
Why moving toward AI systems feels less like a reset and more like an extension of production software engineering.
AI Lab
A public notebook for prototypes, evaluations, and questions.
Ideas become small systems, tests, and better questions. Each entry documents the architecture, evidence, and limitations.
Enter the labLAB 01Planned prototype
Grounded answers with retrieval
How do chunking, retrieval, and citations affect the usefulness and traceability of an LLM answer?
RAG · retrieval quality · grounded evaluationLAB 02System sketch
An agent loop that fails visibly
How should tools, state, validation, and recovery fit into a dependable agent workflow?
Tool use · state · validation · recovery
Who I am
I’m drawn to the space between disciplines.
Mechanical engineering taught me to see systems. Robotics and space widened the scale. Software taught me to build for the real world. AI is opening new questions about intelligence and people.
Meet me beyond the résumé