Ashish Vaswani

Ashish Vaswani grew up in India, earned his B.Tech at the Birla Institute of Technology, and then made his way to the University of Southern California for his PhD in computer science. He wasn't the most prominent name in the field when, in 2017, he sat down with seven Google Brain colleagues to write what would become one of the most cited papers in the history of computing. The question consuming him was simple but radical: what if instead of processing a sentence word by word, the way networks of the era did it, a machine could look at the entire sequence at once and decide which parts deserved more attention?

The answer came in June 2017 with "Attention Is All You Need," presented at NeurIPS that same year. The paper proposed the Transformer architecture, a system that replaced the recurrent networks then dominating the field with an attention mechanism that lets a model weigh how every word relates to every other word simultaneously. This wasn't an incremental improvement, it was a complete rupture. BERT, GPT-3, ChatGPT, Claude, Gemini — all of them were born from that idea. The paper has accumulated more than 200,000 citations, one of the highest counts in the history of computer science.

After his work at Google Brain, Vaswani co-founded Adept AI in 2021 alongside other colleagues, though he left the company in less than a year to pursue a different vision. In 2023, together with fellow Transformer co-author Niki Parmar, he founded Essential AI, a company focused on AI agents for enterprise workflows. Essential AI raised $56.5 million in its Series A in December 2023 and reached a $1 billion valuation with its Series B in 2025. Vaswani is now its CEO.

What strikes me most about Vaswani's story is how quiet it has been. No shows, no polarizing declarations, no public drama. Just a paper, written with a team, that quietly ended up underneath everything. Every time you write to Claude, every time ChatGPT answers you, every time a translator makes sense of your sentence, there's an architecture that traces directly back to his idea: that paying attention, really paying attention, to everything at once, changes everything. Not a bad lesson for those of us who use AI every day.