This is not a startup story. It is a longer story than that.
I came to Chennai to study Visual Communication.
I did not get in.
That is how it starts. Not with a plan. With a door that did not open.
I ended up at Loyola College instead, in the English Literature department. My father was a Tamil Professor — he might have approved of literature, even if it was the wrong language. I was eighteen, in a city I had moved to for one dream, sitting in classrooms built around another.
My father died in my second year.
He was our only income. Overnight, Chennai became a city where I needed to survive before I could do anything else.
The midnight shift
So I worked. Evenings at a xerox and STD shop. Midnight shifts at a browsing center. The kind of place that stays open late because the people who need it most cannot afford to come during the day.
I stopped going to my classes. I did not make a speech about it. I just stopped.
And then, because films had always been the thing I actually wanted, I walked into Prasad Studios as an apprentice editor. One of India’s great film institutions. I learned cinema from the inside — how a story is assembled, how a cut works, what an audience feels in their body when time is manipulated correctly. I learned that storytelling is not a literary skill. It is an experiential one.
Over the next twenty years, I produced television that millions of people watched, and films that travelled to Busan and back. I learned, in the most direct way possible, how human beings absorb things they did not know they were going to remember.
The dream that did not change
Through all of it — the midnight shifts, the studio floors, the television sets, the festival circuits — I carried one idea that never left me.
All learning should be interactive, immersive, and experiential. Without removing the teacher. Without replacing the classroom. Not instead of school. Inside school.
I started building Karkei with that idea. Grade 6 to 12, STEM subjects. I knew it would work for history and geography. I knew younger children would come alive inside it. I also knew the full content for even one grade would take years to build. And a startup cannot survive on years.
I needed a bridge.
The car
My car got damaged. I took it to a body shop in Chennai.
While I was waiting, watching the painter work — the spray gun, the distance from the surface, the angle, the overlap between each pass — I felt something. Not about the paint. About the painter. The way no one had ever taught him this except another painter watching him do it wrong until he got it right. The material wasted in each mistake. The years it took to build a skill that a better system could teach in weeks.
I started researching.
Spray painting led me to welding. Welding led me to dental training. Dental training led me to the numbers: dental colleges in India spend INR 40 to 80 lakhs per station on physical phantom heads that give the trainee zero objective feedback. No score. No record. No way to know that your instrument angle was off by twelve degrees on the third pass.
Skill training — vocational, technical, procedural — is a declared national priority. Millions of trainees every year. Methodology unchanged for decades. The gap was not small.
What filmmaking taught me about training
Every existing training simulator makes one mistake I could see the moment I encountered them.
They put a headset on the trainee.
From a filmmaking perspective — twenty years of understanding how human beings experience things together — this is the wrong answer. A VR headset isolates you inside a private screen. The instructor cannot see what you see. The student beside you cannot watch what you are doing. You cannot learn from each other. You cannot be corrected in shared space.
Cinema works because everyone in the room experiences the same thing at the same moment. That shared presence is not incidental — it is the mechanism. The story lands differently when other people are feeling it with you.
Skill training works the same way.
So I designed a system where the training environment is projected onto a surface that everyone sees. The trainee holds the real tool — the actual spray gun, the actual welding torch, the actual dental handpiece — fitted with a small sensor assembly that tracks every movement in real time. The simulation runs on the projected surface. The instructor watches live. Every student in the room watches live.
No headset. Shared space. Real tool. Objective score.
In April 2026, we ran a pilot at The Rajavva Public School in Chennai. It worked. Every defect type the system was designed to detect, it detected. Every score was reproducible. The instructor watched the trainee’s technique in real time without interrupting the session.
On 29 May 2026, I signed a provisional patent specification as its sole inventor. Filed with the Indian Patent Office. Application number: 202641067392.
The next day
The day after I filed the first patent, I filed the second.
If the first patent was the bridge — the product that could generate revenue while the original dream was being built — the second patent is the dream itself.
It is called the Universal Learning OS. A phase-enforced, gesture-controlled, shared spatial learning system. No device per child. No headset. No internet required. A projector, a depth-sensing camera, a physical surface — and a pedagogical framework built into the system itself, not left to teacher memory or curriculum documentation. The system enforces the learning sequence as a technical constraint. You cannot skip the exploration phase to get to the assessment. The brain’s actual conditions for retention are not optional settings. They are architecture.
Subject-agnostic. Setting-agnostic. School classroom or Anganwadi centre or village community hall or industry training room. Grade 6 Biology or rural health awareness or pre-primary early learning. The same hardware. Different parameter sets loaded at the start of each session.
This is the idea I have carried since I was eighteen years old, working midnight shifts in a browsing center, thinking about what learning could be if it was designed the way experience is designed.
30 May 2026. Application number: 202641067862.
What Angel goes to school for
I have three daughters. My eldest is Angel. She goes to school every morning inside a system I have spent years trying to understand and, now, years trying to improve.
I watch her study. I watch her forget. I watch her come alive in moments that have nothing to do with what is being examined. I am not angry about this — I know the system, and I do not think it is run by people who do not care. But I am building Karkei because I believe, from the science and from what we saw in that Chennai classroom in April, that we can do better. Not by replacing what teachers do. By building the conditions that the brain actually needs to hold onto what it encounters.
My father believed the relationship between a teacher and a student was not a transaction. It was something that had to happen in the same room, in real time, between human beings. The whole design of both patents — shared projection, no private headsets, instructor watching alongside the learner — comes from that belief.
I dropped out of the English Literature department he might have wanted me to finish.
But everything I have built is about keeping the teacher in the room.
On 29 and 30 May 2026, I signed two provisional patent specifications as their sole inventor.
The boy who ran the midnight browsing center to survive his second year of college has two patent numbers.
My father would have had something to say about that. In Tamil, probably. And precisely.
— Elango Raghupathy
Founder, Karkei · Chennai · elango@karkei.com
Karkei Experiential Technologies Pvt Ltd (DIPP262419) is a Chennai-based startup building AR-powered, projection-based learning systems for schools, vocational training centres, and community programmes across India. If you are an educator, institution leader, or investor, write to us at vanakkam@karkei.in.