Hire Pratik M. - Generative AI Architect

Transforming Sales Operations through Generative AI Innovation Leading the architecture and development of enterprise-scale generative AI solutions that revolutionize sales operations at NVIDIA. Spearheading multiple mission-critical projects including: Architecting and building SASIE, a production grade RAG based advanced sales literature chatbot enabling AI-powered access to sales content across Nvidia. Featured on LlamaIndex and Nvidia Technical blogs. https://www.llamaindex.ai/blog/case-study-building-an-ai-sales-assistant-with-llamaindex-and-nvidia-nim Developing an agentic Text-to-SQL framework that bridges Databricks, NVIDIA NeMo, and Azure OpenAI, empowering sales teams with natural language access to complex business data Building scalable microservices that integrate generative AI capabilities into Salesforce support systems, significantly reducing ticket resolution time and enhancing support efficiency Expert in large language model deployment and optimization, with hands-on experience managing production workloads on NVIDIA A100s and integrating various LLM architectures (Llama, Mistral) for enterprise use cases. Skilled in designing RAG-based systems, Agentic workflows, vector databases (Milvus, Pinecone), and enterprise knowledge management solutions. Passionate about creating practical AI solutions that drive business value while maintaining enterprise-grade security, scalability, and performance standards. Consistently collaborating across teams to deliver innovative solutions that enhance sales productivity and customer experience.

I served as the Head of AI in a cross-functional role, where I led a diverse team of engineers, PhDs, and data scientists, fostering a culture of innovation and excellence. I not only provided technical leadership but actively engaged in hands-on development. My responsibilities also included direct interactions with customers, investors and the c-level executives. In my day-to-day work, I was deeply involved in the entire pipeline of full stack machine learning from development to deployment of custom AI solutions. I pioneered various NLP techniques, including LLMs (medical coding), transformers (quality reporting), and GPT (disease progression). A preprint of my latest paper using LLM is available on arXiv (https://arxiv.org/abs/2310.04595) I also led the development of a comprehensive healthcare platform and real-time insights tools, scaling to 40 million individuals, leading to proactive interventions for high-risk patients and cost savings of $50 million.

Developed a drug-drug interaction app - orchestrated with streamlit, which detects drug interactions for a molecule with 90% accuracy. Leveraged keras, python, streamlit and AWS to build an end to end ML pipeline with a custom ’CNN-LSTM with attention’ architecture.