My Professional Journey

Quickcall.dev Founder (Dec 2025 - Present) Solving for developer productivity by solving for interruptions in their workflow Building QuickCall, an AI assistant that helps you connect your org tools and workflows without switching context Yral AI Consultant (Jan 2025 - Nov 2025) Built RAG based content moderation system (86.8% accuracy, +26%) using Phi-3.5 4B Deployed content moderation system on T4x2 GPU (19K req/USD, 3 req/sec) Optimized LLM inference with KV caching: ~70% cache hit rates and ~150 tokens/sec throughput on T4 GPU Developed real-time recommendation system with online learning capabilities, architected to support million-scale user interactions Architected end-to-end candidate generation and personalization pipeline with Airflow, GCP BigQuery, Redis caching Self-hosted MuseTalk talking head for video generation, supporting both image-to-video and video-to-video avatar synthesis People+AI AI Consultant (Nov 2024 - Dec 2024)