From MBBS to AI in a Matter of Months

Dr. Nikhil Rao* came to us with a finished medical degree and an unfinished question. He had his MBBS behind him, but what pulled at him now lay in a different direction entirely — computational neuroscience and machine learning. He wanted to understand how ML actually works. There was only one problem, and he named it plainly at his first meeting: his training was in medicine, not mathematics. The last time he had studied math was in the twelfth grade, and that was many years ago. It is the kind of gap that talks people out of starting.

His mentor did not let it. She reframed the whole thing. This would be swift, she told him, because he had already learned these things once — and the brain does not truly lose what it has learned. It stores it, and it can be revived, often coming back sharper than before, with a fresh perspective and a better understanding. He would be surprised, she promised, how quickly it returned. She was right. Within a few weeks Nikhil found he could move into the material with real ease, and even where a topic was genuinely hard, the ongoing discussions carried him through it quickly. With his mathematical footing restored, the two turned to neural networks — and now the ideas landed. Back-propagation, loss functions, the concepts that sit at the foundation of how a network learns, all of it assimilated smoothly because the groundwork was solid beneath him. In two months Nikhil learned Python from scratch and was building basic neural networks with his own hands, from the ground up.

But the technical craft was only half of what was happening. Because Nikhil kept circling back to the brain itself — its structure, and how it gives rise to cognition, learning, and intelligence — his mentor pointed him to Jeff Hawkins's On Intelligence. He found it captivating, and it left him with a bigger question: could the brain be changed, rewired, in a way that might treat neurodegenerative disease? He went looking for the answer in David Eagleman's Livewired, a deep exploration of a brain that is never fixed but always rewiring itself. Almost without noticing, he was reading widely and hungrily — Ray Kurzweil's The Singularity Is Near, Ramachandran's Phantoms in the Brain, Barbara Arrowsmith-Young's The Woman Who Changed Her Brain, and more. For those months he was quietly absorbing, sharing only fragments of his thinking with his mentor. But she could see it clearly through those glimpses: ideas were forming, and Nikhil would soon have real work to do.

Then the dots connected into a single insight. He would use the power of AI and machine learning for the early diagnosis of neurodegenerative disease — Parkinson's above all. And the confidence that came with that clarity was earned: a man who had started completely from scratch had, in three or four months, rebuilt the fundamentals of mathematics and machine learning and gained a genuine grip not just on the topics, but on his purpose. From there he threw himself into research. He learned the specifics of Parkinson's, the latest developments in diagnosing it, and confronted a hard truth at the center of the problem: by the time conventional methods can detect the disease, much of the damage in the brain has already been done. That was the gap worth closing. Over the next four months, with only minimal guidance from his mentor, Nikhil built a machine learning algorithm that could detect patterns in the voices of Parkinson's patients years before traditional tools would ever raise a flag. A remarkable achievement — the kind of early warning that could change what treatment is even able to do.

He did not stop at the result. With help from other mentors at the Centre, Nikhil connected with a wider community working toward the same goal and founded a startup around his work, resolved to turn technological advancement into real progress for medical science. His story carries a message for every person who believes they have started too late or come from the wrong background. Nikhil had a medical degree and no recent mathematics, and in a matter of months he was writing algorithms that reach where established medicine cannot yet see.

* A pseudonym — names have been changed to protect the privacy of our students.