Anurag Singh

Spearheaded the design, development, and launch of VocRT, a real-time, privacy-first conversational voice AI system—built entirely from the ground up as a solo open-source project. Engineered a fully local processing stack integrating Whisper (STT), Silero VAD, Kokoro TTS, and Retrieval-Augmented Generation (RAG) with Qdrant. Achieved ultra-low latency, seamless interruption handling, and API-free operation. VocRT empowers secure, natural voice interactions across domains like accessibility, education, healthcare, and AI-driven customer support.

During my internship at IVY, I contributed to the development of a Voice AI-based web application. My role involved end-to-end web development using technologies such as Firebase DB, Express.js, React.js, and Node.js. I also worked on integrating third-party SDKs & APIs, DevOps tasks, and managing GCP infrastructure. I quickly learned and implemented WebRTC to build real-time communication features and efficiently integrated Large Language Models (LLMs) into the application. My proactive approach and ability to adapt to new challenges were recognized by the team, culminating in the successful delivery of the project.

Designed and implemented a comprehensive Database Dashboard to streamline Data Visualization. Built an AI-powered meeting booking system that automates scheduling and sends meeting invitations directly via Gmail. Leveraged AI to optimize the booking process, enhancing user experience and efficiency.