The story of artificial intelligence is often associated with the biggest technology companies and the most prestigious universities in the world. But behind some of the research that helped shape today’s AI revolution are engineers whose journeys began far from those institutions. Niki Parmar, an Indian computer scientist and AI researcher, is one such name. A graduate of Pune Institute of Computer Technology, Parmar went on to become part of a Google Brain research team whose work would have a lasting impact on modern AI.Her journey took her from an engineering college in Pune to the University of Southern California, followed by professional stints at Amazon and PubMatic and, eventually, Google’s AI research ecosystem. At Google, she became one of the eight researchers behind “Attention Is All You Need”, the 2017 paper that introduced the Transformer architecture and became a foundational piece of modern large language models. The paper predates ChatGPT, Claude and Gemini, but its ideas subsequently became central to the technology powering today’s generative AI systems.
From PICT to the global AI research community
Parmar completed her engineering at Pune Institute of Computer Technology before moving to the US for higher studies. She earned a master’s degree in Computer Science from the University of Southern California and subsequently gained industry experience through an internship at Amazon and work at PubMatic.Her transition into AI research eventually took her to Google, where she joined the Google Brain team. It was here that she worked with researchers including Ashish Vaswani on the Transformer research that would fundamentally influence natural language processing.
The research paper that changed AI
Published in 2017, “Attention Is All You Need” proposed a new neural-network architecture based entirely on attention mechanisms. The Transformer offered a more efficient way of processing sequences and subsequently became the foundation for a generation of large language models.Parmar was one of the eight authors of the paper and the only woman in the research group. Profiles of the research team have also noted that she was the only member without a PhD at the time.The importance of the paper became increasingly apparent years later as Transformer-based models moved from research laboratories into products used by millions of people. While Parmar did not build ChatGPT, Claude or Gemini herself, her work on the Transformer architecture formed part of the research lineage that made modern generative AI possible.
From research to entrepreneurship
Parmar later moved beyond academic research to focus on applying AI to real-world products. She co-founded Essential AI with Ashish Vaswani and others. The company raised $56.5 million in a Series A funding round in 2023, with investors including March Capital, Google, Nvidia and AMD.She later joined Anthropic, where she has worked as a Member of Technical Staff.
A career built beyond conventional labels
Parmar’s journey does not suggest that prestigious institutions are irrelevant. Rather, it demonstrates that an engineer’s career can evolve dramatically through skills, research, persistence and the ability to work on problems that matter.From a Pune engineering college to one of the research teams that helped reshape artificial intelligence, her career offers students a powerful reminder: the starting point may be defined by a college, but the direction of a career is ultimately shaped by what one chooses to learn, build and pursue.
