Indian AI startup Sarvam has just pulled off a major coup by bringing AI expert Devendra Chaplot as a part-time advisor from Elon Musk's xAI. Chaplot's arrival comes as Sarvam opens a new office in San Francisco.Devendra Chaplot has joined Sarvam, leaving Elon Musk's xAI. (Photos: Sarvam, X/@@dchaplot)The Indian AI startup Sarvam has just bagged a big win in the AI talent wars. Devendra Singh Chaplot, a star AI researcher, has joined Sarvam, leaving behind Elon Musk’s xAI (now SpaceXAI). He will be working as a part-time advisor at the AI startup.In case you are unaware, Devendra Chaplot is one of the high-profile AI researchers in Silicon Valley. Before joining Elon Musk and xAI, he was part of founding teams of two different AI startups – France’s Mistral, and former OpenAI executive Mira Murati’s Thinking Machine Labs.Devendra Chaplot’s arrival could be a big boost for Sarvam, which is now building a trillion-parameter AI model in India from scratch. It is also the first time that an Indian startup has managed to bring in a key AI researcher at a time when companies like OpenAI and Anthropic are trying to poach the best talent by paying millions of dollars.Chaplot appeared on stage during Sarvam Epoch 2026 in Bengaluru. He expressed his excitement in joining the Indian startup, but highlighted that Sarvam needed key talent. "I have been part of founding teams at frontier labs. You need a handful of experts and experienced people but we need more of motivated talent, who can catch up very quickly and run the company," he said. Devendra Chaplot’s arrival comes at a time when Sarvam is opening a foundational AI research lab in San Francisco, in the US. Chaplot, who is based in Palo Alto, California, will be working from this lab.Who is Devendra Chaplot?Chaplot is an IIT Bombay alumnus who has spent over a decade in the field of AI. In March this year, he joined Elon Musk’s xAI. But as per his LinkedIn profile, Devendra Chaplot left the startup in April after a brief stint. Before xAI, the AI expert was the tech lead for data and pretraining at Thinking Machine Labs – the AI startup founded by Mira Murati and other former OpenAI executives. Here he worked on the Tinker training API.Devendra Chaplot was a founding researcher at Mistral AI, where his work included training Mistral 7B, Mixtral 8x7B and Mistral Large, and helping set up Mistral's US office in Palo Alto.Earlier, Chaplot worked at Facebook AI Research and had also spent time as a research associate at Apple and Samsung. His research has focused on machine learning, robotics, computer vision, embodied AI, navigation, multimodal systems and large language models.In terms of qualifications, Devendra Chaplot holds a PhD in Machine Learning from Carnegie Mellon University, where he also completed his Master's in Language Technology. He has done his BTech in Computer Science and Engineering from IIT Bombay.Devendra Chaplot says Sarvam needs motivated talentGoing forward, Devendra Chaplot insisted that there was a need for Indian AI models that could meet local requirements. He added, “For India, we have so many languages and cultures. We want to have models that can be customised. This is the reason we absolutely need to build models here.”Though Chaplot accepted that models can become outdated over time, the learnings that come only help build more in the future. “Model is not the goal, it becomes obsolete, but knowing how to build and train your models will outlast everything else,” the Sarvam advisor said.Devendra Chaplot’s expertise will come in handy for Sarvam, which is rapidly expanding. At Sarvam Epoch, the AI startup made a bunch of new announcements, including Saras V4, Bulbul V4 and Vision 2.0Sarvam is India’s first AI unicorn with a $1.5 billion market cap after raiing $234 million in Series B round led by HCLTech.- EndsPublished By: Armaan AgarwalPublished On: Jul 30, 2026 13:35 IST
AI expert Devendra Chaplot was working with Elon Musk, Sarvam steals him from xAI
Full Article
Original Source
Read the full article at Indiatoday →KhanList aggregates and links to publicly available news content. We do not host full articles from third-party sources. Always verify important information with original sources.