Voice AI & Telephony Engineering
LiveKit (WebRTC), SIP / VoIP bridging, WebSockets, Gemini Live API, Sarvam, Ultravox, Groq, streaming Speech-to-Text (STT) and Text-to-Speech (TTS) low-latency pipelines.
Categorized Technical Capabilities
AI Orchestration & RAG Frameworks
Google Agent Development Kit (ADK), LangGraph, Multi-Agent Systems, A2A Protocol, Model Context Protocol (MCP), Retrieval-Augmented Generation (RAG), LiteLLM, LightRAG, Vector DBs (ChromaDB, Qdrant, Pinecone).
Evaluation & Quality
LLM-as-a-judge scoring, human review workflows, span-level error annotation, word-level language-confusion tracking for multilingual (Indic) voice calls, agent test playgrounds.
Generative Media & Web Automation
Google Veo (video generation), Imagen (image generation), Cartesia (primary TTS and voice cloning in VOAG), Fish Audio (voice cloning), FFmpeg video stitching, Playwright (headless scraping & visual brand extraction).
Backend & Concurrency
Python (`asyncio`, `contextvars` session metering), Go (`whatsmeow` session library), SQL (relational schema design), JavaScript / Next.js (front ends for AI apps), FastAPI, REST APIs, OAuth 2.0, Redis, NGINX, Docker, Microservices Architecture.
Cloud & Infrastructure
Google Cloud Platform (Cloud Run, Compute Engine, least-privilege IAM, VM provisioning & hardening, cost control), DigitalOcean, ZeroSSL, Git, GitHub Actions (CI/CD), NAT Gateways.
Machine Learning & Data Science
SARIMA, XGBoost, ARIMA, Prophet, Predictive Modeling, Time-Series Forecasting, Computer Vision Inference.