Speaker Profile
Ritesh Vajariya
Founder and CEO,Ritesh Vajariya is Founder and CEO of AI Guru®, a practitioner-led venture studio that builds vertical AI products, runs enterprise AI training, and advises boards on AI governance.
He is one of the few practitioners who has architected frontier AI at the infrastructure level and ships production systems today. At Cerebras Systems, he led GenAI strategy on wafer-scale compute. At AWS, he led the team that partnered with Bloomberg on BloombergGPT, one of the most sophisticated enterprise large language models ever deployed; he also owned Amazon SageMaker and contributed to the Bedrock launch, helping grow AWS AI revenue from $40M to over $700M. Before AWS, he spent five years on Bloomberg's machine learning platform team.
Today he builds on Claude. His industrial AI platform, MillMind™, runs live in a working paper mill, an AI operating layer on the Claude API deployed where downtime carries real cost. He has used Claude Code to run a multi-phase, multi-agent security review across his own production stack before third-party penetration testing. This is the lens he brings to teaching: what holds up when an agent is doing real work against real systems, not a demo. He has trained 100,000+ professionals across five continents and shipped 20 deployment-ready AI products on the Claude stack.
He is one of the few practitioners who has architected frontier AI at the infrastructure level and ships production systems today. At Cerebras Systems, he led GenAI strategy on wafer-scale compute. At AWS, he led the team that partnered with Bloomberg on BloombergGPT, one of the most sophisticated enterprise large language models ever deployed; he also owned Amazon SageMaker and contributed to the Bedrock launch, helping grow AWS AI revenue from $40M to over $700M. Before AWS, he spent five years on Bloomberg's machine learning platform team.
Today he builds on Claude. His industrial AI platform, MillMind™, runs live in a working paper mill, an AI operating layer on the Claude API deployed where downtime carries real cost. He has used Claude Code to run a multi-phase, multi-agent security review across his own production stack before third-party penetration testing. This is the lens he brings to teaching: what holds up when an agent is doing real work against real systems, not a demo. He has trained 100,000+ professionals across five continents and shipped 20 deployment-ready AI products on the Claude stack.