I'm an AI engineer based in Islamabad, currently running AI GenMat as Managing Director while finishing my BS in Artificial Intelligence at FAST-NUCES Peshawar (2025 to 2029). Most of my work sits in agentic AI, RAG pipelines, NLP, and computer vision, basically anything where an LLM needs to actually do something useful instead of just chat.
name: Muhammad Awais
role: AI Engineer, Managing Director @ AI GenMat
location: Islamabad, Pakistan
education: BS Artificial Intelligence, FAST-NUCES Peshawar (2025-2029)
contact: m.awais@aigenmat.com | mawaisqq@gmail.com
focus:
- Agentic AI and multi-agent systems
- LLM fine-tuning and RAG pipelines
- NLP and computer vision
- GenAI & Automation
languages:
- Python
- C++
- C
- Bash
open to:
- High impact AI collaborations
- Freelance AI systems and infrastructure work
- Research oriented AI engineering
- Agentic AI consulting
portfolio: https://ik-awais.github.io|
A full stack AI research engine. It pulls from multiple search providers (Tavily, Brave, Firecrawl), runs RAG over your own documents with pgvector, and streams cited answers back through a Next.js frontend. JWT auth included, everything runs in Docker Compose. |
A RAG study assistant built for actual coursework. Upload PDFs, DOCX, or PPTX lecture slides, ask questions in plain English, and get back answers with the exact page cited. Indexing runs on a non blocking thread pool, files get deduped by SHA-256, and chat history is session aware. |
|
An AI companion that picks up on mood. It uses VADER for real time emotion detection, drives a Three.js orb that morphs as your mood shifts, and adjusts its tone for PK, IN, AE, US, and GB users. Memory persists in SQLite, and it can switch between Gemini 1.5 Pro and NVIDIA NIM depending on the request. |
A desktop toolkit for everything file related. Convert, split, merge, protect, and watermark PDFs, DOCX, XLSX, PPTX, CSV, images, audio, and video, all from one app. There's a built in Gemini assistant for asking questions about your documents and an NVIDIA powered image upscaler. No cloud needed. |



