VOCALIQ — AI Voice Customer Experience Platform
Real-time bilingual (Urdu and English) voice AI platform with multi-agent workflows, full-duplex WebRTC voice calls, barge-in interruption handling, and Graph RAG knowledge grounding.
Applied AI Engineer
I design and build production-grade AI systems: LLM applications, RAG pipelines, and autonomous agents that solve real business problems.
Applied AI Engineer with 3+ years of experience across ML engineering, Generative AI, and MLOps. I am currently an AI Engineer at Excels Tech Solution LLC in Lahore, Pakistan, and previously worked as an ML Engineer at PerceptronAI.
I take AI products from prototype to production: APIs, vector search, agent workflows, observability, and cost and latency optimization. My flagship work is VOCALIQ, a real-time voice AI platform with bilingual Urdu and English agents grounded by Graph RAG.
Alongside industry work, I have also taught at university level as a Junior Lecturer at Superior University. I hold a B.S. in Computer Science from COMSATS University Islamabad, and I am open to freelance projects in LLMs, RAG, AI agents, and AI automation.
The tools I use to design, build, and run AI systems in production.
Open-source AI systems I designed and built end to end, from backend architecture to the user interface.
Real-time bilingual (Urdu and English) voice AI platform with multi-agent workflows, full-duplex WebRTC voice calls, barge-in interruption handling, and Graph RAG knowledge grounding.
Upload PDF or TXT documents and ask questions in plain language. Documents are chunked and embedded into ChromaDB, and answers come back grounded with source citations.
An AI agent built with LangGraph that reasons over tools: calculator, live web search, current date and time, and a knowledge lookup, with per-session memory across a CLI, a FastAPI API, and a browser chat UI.
A production-style MCP server exposing four tools: read-only SQL queries, file search, web fetch, and revenue analytics, over stdio and HTTP, plus a web demo client that discovers tools live.
SaaS-style support desk where every incoming ticket is auto-triaged by AI: category, priority, and sentiment are detected, and a reply is drafted before an agent opens it.
One OpenAI-compatible API in front of multiple LLM providers, with API-key auth, per-key rate limiting, retries, streaming, and a live dashboard tracking requests, tokens, cost, and latency.
Building production AI systems with LLMs, RAG pipelines, and agentic workflows.
University-level teaching alongside industry engineering work.
Machine learning engineering: model development, data pipelines, and ML-backed product features.
Bachelor's degree in Computer Science.