Varshit Manepalli

Background

About Varshit Manepalli

Machine Learning Engineer with over 3 years of experience delivering scalable, intelligent systems across domains such as healthcare, cybersecurity, and natural language processing. Adept at developing and deploying end-to-end ML pipelines from data preprocessing and model training to API integration and production deployment using tools like TensorFlow, PyTorch, Flask, and GCP Vertex AI. Currently contributing to Lucille, an AI-driven mental health platform, where I’ve led the development of emotion detection systems using facial analysis, body language (via BiLSTM and pose estimation on RAVDESS), and environmental context recognition. I integrated speech-to-text and text-to-speech functionality into conversational AI features and built robust recommendation engines based on real-time mood tracking, user history, and behavioral patterns. Previously at Progress Solutions, I engineered a BERT-based hate speech detection system and optimized LLM-driven chatbots, improving user engagement and minimizing model bias. My contributions to cybersecurity include designing a GAN-based threat simulator generating 200k+ synthetic attacks to preemptively detect malicious patterns. I bring cross-functional experience, having collaborated with UX researchers, DevOps engineers, and clinical professionals to build user-centric, data-driven applications. With deep technical skills in Python, SQL, REST APIs, and cloud ML platforms, I thrive in fast-paced environments where innovation, impact, and reliability matter most.