Vishal Agarwal

Founded and led a multi-agent AI platform for biotech with multimodal data capture and voice/conversational workflow execution. Implemented modular microservices using Docker, EC2, and S3 for high-throughput biotech data processing. Deployed a scalable system with a Supabase backend enabling real-time collaboration and enterprise-grade data management for research teams.

LLM Chatbot: Developed a multimodal face retrieval system and RAG solution for a US defense-backed startup. Implemented MLOps pipelines using Kubernetes, Docker, and AWS for scalable, consistent deployment. AI Lesson Generator: Optimized for cost and latency (70% improvement) and deployed on AWS for a Canada education department vendor.

Oversaw the design of ML-based predictive models for evaluating patient responses in cancer drug trials and identified biomarkers distinguishing responders from non-responders. Led the development of a Deep Neural Network for credit card fraud detection, achieving a 60% improvement over traditional rule-based models; managed dataset optimization to 1/100th of original size.