I'm a full-stack engineer building intelligent developer tools and voice-first experiences. I am currently working on an AI-powered interview platform, where I have integrated LLMs, real-time evaluation services, and multi-agent workflows. Recently, I have been experimenting with speech pipelines—STT, LLM message orchestration, and TTS—to create natural, context-aware conversational systems. I enjoy rapid prototyping, building scalable backend systems (NestJS, Node, Python), and designing production-ready AI features.

Advanced TTS generation and emotion control – exploring nuanced delivery, prosody, and contextual adaptation. Streaming STT + LLM orchestration – building low-latency conversational agents with smart memory, phase transitions, and multi-agent coordination. Agentic voice interfaces – creating voice-driven tools that can reason, plan, and execute tasks in the background. Voice UI/UX design – understanding what makes spoken interactions intuitive, frictionless, and “human-feeling.”

1. Real-Time Voice Agent Pipeline (STT → LLM → TTS): Building a streaming voice interaction system that handles user speech, routes it through an LLM, and returns natural TTS responses. Experimenting with phase-based transitions, instruction updates, and triggering follow-up LLM calls after TTS playback. 2. Multi-Agent Interviewer for Coding Assessments: Developing an AI interviewer that can ask technical questions, evaluate responses, and adapt its behavior based on candidate performance.

Founding CTO of Glints, one of the biggest job platforms for young people. Now working on machine learning applications for the marketplace.

Voice AI, agents in real world applications, would love to hear about boots-on-the-ground contact

Working on a story telling tool for my friend to read to his kid