A young founder-engineer, still early, already building.
I build at the intersection of AI, hardware and real-world systems. I care about taking ambitious technical ideas from a rough concept to something that actually runs, and being honest about what worked and what didn't.
ideas → systems → impact
Divyanshu Dubey
Bhopal, India
I'm Divyanshu, a computer science student and founder based in Bhopal, India, building globally. Most of what I've learned has come from building things, putting them in front of people, and watching where reality disagreed with my assumptions.
My strongest work sits where software meets the physical world: Enfacial, a privacy-first, face-authenticated payment system for campuses, and Ocuvision, AI-powered smart glasses combining embedded hardware, computer vision and software.
I'm less interested in shipping demos and more interested in whether an idea should exist at all, which means caring about users, distribution and the unglamorous parts as seriously as the technology.
I'm still early. The point of this site isn't a list of achievements. It's the trajectory.
Problems I keep coming back to.
Human-computer interaction
How can computers understand context without constantly demanding our attention?
Intelligence in the physical world
What happens when AI moves beyond browsers and screens?
Identity without friction
Can authentication become invisible without sacrificing privacy or security?
New computing interfaces
What comes after the smartphone-centric model of computing?
How I build.
Problem
Find something people actually struggle with.
Hypothesis
Develop a strong opinion about how it could work differently.
Prototype
Build the smallest system capable of testing that opinion.
Users
Put it in front of real people quickly.
Learn
Understand what was wrong.
Iterate
Improve, reposition, or stop.
“I care more about discovering whether something should exist than proving that my first idea was correct.”
Things that didn't work.
Building has changed my mind about a lot of things.
Building before distribution
Lesson
A technically strong product doesn't automatically create demand. Distribution has to be designed as seriously as the product.
Institutional adoption
Lesson
Better technology doesn't remove trust, procurement and deployment friction.
Building too broadly
Lesson
A product usually needs one obvious reason to exist before it needs ten features.
Projects that stopped
Lesson
Stopping something can be the correct outcome of an experiment.
Still early.
A timeline of things I've tried, built and learned.
Started seriously building software and experimenting with products.
Worked on products and startup experiments.
Started building increasingly ambitious systems across AI, hardware and real-world applications.
Focused increasingly on intelligent systems, distribution, and finding problems worth spending years on.
Expected B.Tech Computer Science graduation.
still building.
I can build the thing.
Across product, AI, systems and infrastructure, enough to take an idea from concept to a working system.
Product
AI
Systems
Infrastructure
Languages
Building
Intelligent products and experimenting with new startup ideas.
Thinking about
- AI × hardware
- Human-computer interaction
- New computing interfaces
- Distribution
- Education
- Identity
Looking to meet
- Exceptional founders
- Technical builders
- Researchers
- Startup operators
- Investors who enjoy talking with builders early
Also open to unusually high-agency roles where I can build close to founders, own large technical problems, and move quickly.