AI ENGINEER & INTERDISCIPLINARY RESEARCHER

Ideas into
systems.
Evidence into
intelligence.

I’m Muhammad Abdur Rafey. I build software that knows where its answers come from.

NOWEngineering agentic products at Quest

GitHub ↗

Multi-agent systemsKnowledge graphsRetrievalParallel computing

SELECTED WORK / 01 — 07

Built to solve
something real.

CLIENT SYSTEMS · NDA

One conversation. Many sources of truth.

An architecture that retrieves data at runtime across separate platforms, bringing reports, analytics and agent workflows together without copying or altering the underlying source data.

Multi-agentKnowledge graphsSQL + APIs

Discuss the architecture ↗

RESEARCH / CONNECTING THE EVIDENCE

Trust is an
architecture.

Which source should a system trust, and when?

LIVE SOURCES → REASONING

Cross-verification

Coordinate retrieval and compare evidence before forming a response.

Knowledge graphs × LLMsGrounding agents in structure
LLMs for parallel codeLoops that parallelize themselves
Hybrid quantum–classicalCircuits that seed search
Low-resource speechUrdu text to voice

JOURNEY / ALWAYS CONNECTING DOTS

Breadth,
with a direction.

2021–25 / CHAPTER 01

FAST-NUCES

BS Computer Science

Foundations first: C++, data structures, assembly, databases and competitive programming.

AddingC++Data structures8086 assemblyDatabases & SQLCompetitive programming

KNOWLEDGE BUILDS ON WHAT CAME BEFORE0 nodes · growing

OPEN TO AI ENGINEERING & RESEARCH COLLABORATIONS

Your next
hard problem.
Let’s talk.

AI engineering, research, or an ambitious idea that needs a collaborator.

m.a.rafey1215@gmail.com ↗

LinkedIn ↗GitHub ↗

ALONG THE WAY

1st · ICPC preliminary round
1st · NUMLogic speed programming
3rd · NUST Comppec
MIT iQuHACK 2026 · NVIDIA challenge

Ideas intosystems.Evidence intointelligence.

Ideas into
systems.
Evidence into
intelligence.