When Bengaluru-based Sarvam AI unveiled its large language models at the India AI Impact Summit 2026 in New Delhi, the response was electric. The startup's pavilion drew record crowds at Bharat Mandapam, with thousands of students, developers and professionals queuing to witness live demonstrations - and organisers reportedly extended the expo by a day. The 105B model, built on a mixture-of-experts architecture with a 128,000-token context window, outperforms DeepSeek R1 and Google Gemini Flash on several benchmarks - a jaw-dropping achievement for a startup that barely existed three years ago.
Behind every benchmark and every demo are real people. Here are the engineers who gave blood, sweat, and code to make it happen. These engineers are making India proud.
1. Sumanth Doddapaneni: ML Researcher
A PhD student at IIT Madras currently on leave, Sumanth holds a BTech from IIIT SriCity and is a Google PhD Fellow (2023). His research on multilingual language modelling has earned Outstanding Paper Awards at both ACL 2024 and EMNLP 2024. He previously interned at Google Research in Bengaluru and Mountain View, and at Mila — Quebec AI Institute.
2. Mohit Singla: Founding Member
An IIT Madras alumnus, Mohit brings rigorous research training from one of India's top engineering institutions to Sarvam's model development work, contributing to the language understanding capabilities that make the LLMs genuinely useful for Indian users.
Mohit Singla, ML Engineer at Sarvam AI |
3. Aashay Sachdeva: Founding Team, Model Training
A VIT graduate, Aashay is one of Sarvam's founding engineers with five years of industry experience across healthcare AI at Stasis and real-time ML systems at Mobile Premier League, where he rose from Data Scientist to Senior Data Scientist. He holds a patent in gaming-sector matchmaking - an indicator of the kind of creative problem-solving he brings to Sarvam.
3. Aashay Sachdeva, Founding Team, Model Training at Sarvam AI |
4. Anand Bollu: ML Engineer
A key contributor to Sarvam's model architecture team, Anand works on the systems-level engineering that makes training and deploying frontier-scale Indian language models practically feasible. He has previously worked with Applied Institution in California and is an alumni of the Carnegie Mellon University.
5. Harveen Singh Chadha: LLM Researcher
Harveen left Microsoft's Speech Team, where he built end-to-end speech recognition models, to join Sarvam roughly ten months ago - a move his parents didn't initially support. His LinkedIn bio read 'Building4India' even while at Microsoft. After the AI Summit spotlight, his mother was reportedly promoting Sarvam in neighbourhood WhatsApp groups. He studied at University College of Engineering, Punjabi University.
5. Harveen Singh Chadha: LLM Researcher at Sarvam AI |
6. Arpit Dwivedi: ML Engineer
An IIT Kharagpur graduate, Arpit previously worked as a Data and Applied Scientist before joining Sarvam. His systems background strengthens the team's ability to build efficient, production-grade AI models.
Arpit Dwivedi: ML Engineer at Sarvam AI |
7. Tanay Anand: ML Engineer
An IIT Kanpur alumnus, Tanay previously worked as a researcher at Adobe's Media and Data Science Research division. He brings experience from one of India's premier creative-tech environments to Sarvam's LLM training pipeline.
Tanay Anand: ML Engineer at Sarvam AI |
8. Sushil Khyalia: ML Researcher
Sushil holds a BTech from IIT Bombay and a Master's from Carnegie Mellon University - one of the strongest dual-pedigree combinations in the team. Before Sarvam, he was an engineer at Samsung Electronics, giving him rare exposure to both frontier research and large-scale industrial deployment.
Sushil Khyalia: ML Researcher at Sarvam AI |
9. Manav Singhal: ML Engineer
Manav came to Sarvam from Microsoft Research India, where he evaluated LLMs to automate coding tasks. At Sarvam, he focuses on scaling reinforcement learning. An NIT Karnataka graduate, he was also selected to attend Google Research India's Research Week - a mark of serious academic standing.
Manav Singhal: ML Engineer at Sarvam AI |
10. Raghavan AK: ML Engineer
Raghavan brings a diverse career trajectory to the team - from developer at ThoughtWorks, to ML Engineer at Indix, to Senior Data Specialist at Ericsson - before landing at Sarvam. An IIT Madras alumnus, he exemplifies the kind of industry-hardened engineer Sarvam has assembled alongside its academic talent.
Raghavan AK, ML Engineer at Sarvam |
11. Harshit Kedia: ML Engineer
A Dual Degree Computer Science graduate from IIT Madras, Harshit represents the next wave of IIT talent choosing the startup path. His systems-level CS training feeds directly into the low-level model optimisation work that makes Sarvam's benchmarks possible.
Harshit Kedia: ML Engineer at Sarvam AI |
12. Ritvik Aryan Kalra: AI Engineer
Ritvik holds a dual degree in Computer Science from IIIT Hyderabad, where he also conducted undergraduate AI research. His academic grounding in applied AI makes him a natural fit for Sarvam's research-forward engineering culture.
Ritvik Aryan Kalra: AI Engineer at Sarvam AI |
13. Anna Upreti: Product, Model Behaviour & Evaluations
A postgraduate in Advanced Computer Science from Ashoka University, Anna brought a rare combination of technical rigour and product thinking to Sarvam. Previously at Bain & Company and Delhivery, she worked on model behaviour and evaluations - the unglamorous but critical work of ensuring the models actually do what they claim.
Anna Upreti: Product, Model Behaviour & Evaluations at Sarvam AI |
14. Soham Petkar: Research Fellow
A BTech graduate from Plaksha University, Soham also conducted undergraduate research at Precog at IIIT Hyderabad. One of the younger members of the team, he is part of a generation of builders who are proving that great AI research isn't limited to the traditional IIT pipeline.
Soham Petkar, Research Fellow at Sarvam AI |
This list contains only a few names, and the entire talent pool consists of more than 100 employees. These engineers, most of them in their twenties and early thirties, trained models from scratch on sovereign compute, across 22 Indian languages, with benchmarks that embarrass systems six times their size. They are, quite simply, the people building India's AI future.
