Shubhanjan Shekhar
About Shubhanjan Shekhar
Shubhanjan Shekhar is a Staff Machine Learning Engineer currently working at ThoughtSpot. He has extensive experience in software engineering and machine learning, having previously worked at companies such as IBM, Cisco, and Uber.
Current Role at ThoughtSpot
Shubhanjan Shekhar serves as a Staff Machine Learning Engineer at ThoughtSpot, a position he has held since 2023. In this role, he focuses on machine learning infrastructure and engages in tasks such as converting text to SQL, which is part of his work in generative AI. His expertise includes fine-tuning open-source large language models and utilizing advanced techniques for distributed model training.
Previous Experience at Uber
Before joining ThoughtSpot, Shubhanjan worked at Uber as a Senior Software Engineer in the Uber AI division from 2019 to 2023. During his four years at Uber, he contributed to various projects that involved machine learning and artificial intelligence, enhancing the company's capabilities in these areas.
Experience at IBM and Cisco
Shubhanjan's career includes significant roles at IBM and Cisco. He worked as an Advisory Software Engineer at IBM Watson from 2016 to 2019, based at the IBM Almaden Research Center in San Jose. Prior to that, he was a Software Engineer II at Cisco from 2014 to 2016 in San Jose, California. These positions provided him with a solid foundation in software engineering and machine learning technologies.
Education and Academic Background
Shubhanjan holds a Master of Science (MS) in Computer Science from The Ohio State University, where he studied from 2012 to 2014. He also earned a Bachelor of Technology (B.Tech) in Computer Science and Engineering from The LNM Institute of Information Technology, studying from 2008 to 2012. His foundational education includes high school at Ahlcon Public School and Somerville School in India.
Technical Skills and Expertise
Shubhanjan possesses proficiency in multiple programming languages, including Java, Python, and Go. His technical expertise encompasses distributed model training and serving, utilizing technologies such as Ray and Kubernetes. He employs multi-node multi-GPU distributed training techniques with DeepSpeed and specializes in fine-tuning large language models like Llama and Mistral.