Varad Mhetar

I’m a Speech AI Intern focused on building speech-based AI with modular, testable Python workflows. I developed and integrated components for speech-based AI applications, and I’ve worked end-to-end from ASR through response generation to voice interaction.

In my projects, I built an end-to-end conversational voice AI pipeline that connects speech recognition, LLM inference, and speech synthesis. I designed a modular backend architecture aimed at low-latency, real-time voice interactions, then optimized and debugged audio processing workflows for reliable conversational performance.

I’m also exploring reinforcement learning through an autonomous multi-agent simulation environment. I designed reward mechanisms and environment logic to support reinforcement learning experiments, and I implemented modular components to make experimentation and evaluation extensible.

I’m currently pursuing a B.Tech in Computer Science Engineering (2023–2027), and I bring strong AI/ML fundamentals across NLP, Speech AI, RAG, and LLM orchestration. I’ve been a Google Summer of Code 2024 participant, ranked 2nd in a College Capstone Project, and completed a Hugging Face LLM Fine-Tuning certification.