Multilingual AI, Data Readiness, and Mentorship with Krish Banavalikar

Modernizing with AI is not just about more GPUs or chatbots. In this Winner’s Circle episode, host Russ Fordyce talks with all star judge and Cloud Engagement Manager Krish (Radhakrishna) Banavalikar about what real transformation looks like when you blend data, AI, and human insight. Krish shares his journey from an Indian village to leading global cloud and AI programs, and why he calls himself a “serial learner and unlearner.”​He explains how he helps enterprises move from experiments to value, using agentic AI to design products, shape campaigns, and feed customer feedback back into the loop. The surprising blockers are not usually technology, he notes, but compliance, privacy, and governance that slow deployment and force teams to rethink how they handle customer data. Krish breaks down practical patterns like medallion architectures and data mesh that turn messy operational and transactional data into AI ready assets, and why operational signals often matter more than traditional summaries for understanding real customer behavior.​A highlight of the episode is Krish’s passion for multilingual AI and underrepresented Indo Aryan languages such as Hindi, Marathi, and Sanskrit. He explains why today’s large language models, while seemingly multilingual, still struggle with culture, humor, slang, and mixed language content, and how his own research, including a 97 percent accuracy sentiment analysis project on biased Amazon reviews, points to the need for better datasets and benchmarks beyond English centric standards.​Throughout the conversation, Krish returns to people and careers. He talks about mentoring more than 300 professionals into cloud and AI roles, the impact of early mentors on his own path, and how he tracks job market signals and hype cycles to decide which skills to build for the next three to five years. Russ and Krish close by looking ahead to the massive AI infrastructure build out, the potential shock of quantum computing, and why many enterprises today are simply “keeping up” – caught between old playbooks and a fast moving agentic AI future.​