Iqra Muhammad

PhD ML Researcher | Clinical AI & RAG Architect | Regional Winner @ Ignite

Currently: ML Researcher, Smart EndsExperience: 7 years

PhD-level Machine Learning Researcher specializing in Clinical NLP, Knowledge Graphs, and LLM architectures. I bridge the gap between academic rigor and production-ready medical AI, most notably leading the award-winning MynEdge maternal health platform. With a background in teaching and mentoring 50+ researchers, I focus on building HIPAA-compliant, multi-agent systems that solve high-stakes healthcare challenges. Currently listed with 14 publications in medical AI and NLP venues.

Experience

ML Researcher

Smart EndsBelgium (Remote)part time2025 – Present

Lead a cross-functional ML team delivering end-to-end NLP pipelines, predictive models, and data-driven decision systems for enterprise clients. Evaluate and integrate large language models and transformer-based architectures; mentor junior ML engineers and data scientists.

NLPLLMTransformers

ML Researcher

MynEdge — Maternal AI Platformwork2025 – Present

Lead a multidisciplinary research team building an AI-driven maternal-health platform combining NLP, risk-prediction models, and conversational AI for low-resource clinical settings. Directed the technical strategy that won Regional Winner at the Pakistan Ignite National Technology Fund Championship (2025), a government-backed innovation grant. Own the AI/ML research roadmap — risk-triage models, NLP patient-communication tools, clinical knowledge integration — and coordinate ML engineers, clinical advisors, and developers to move prototypes into deployable product features. Established data-governance and privacy-preserving protocols for sensitive health data; mentor junior researchers and report progress to funders and government stakeholders.

NLPRisk-prediction modelsConversational AI

ML Researcher

PreplyRemoteworkMar 2026

Summary

Advise clients on LLM architecture, RAG pipeline design, multi-agent system design, and production deployment of AI systems. Delivered HIPAA-compliant clinical decision-support and medical NLP solutions for healthcare clients. Ran feasibility assessments, vendor evaluations, and architecture reviews for technical and executive stakeholders.

Results

Redesigned the RAG retrieval strategy for a client's clinical-support system, cutting hallucinated responses by 50%.
Reduced average query latency across a 1000 document knowledge base.

LLMRAGMulti-agent systemsNLP

Skills

Technical

NLP

BERT

Transformers

RAG

Hugging Face

LLMs

Scikit-learn

PyTorch

Python

Knowledge Graphs

Open IE

Multi-agent systems

LangChain

Prompt Engineering

spaCy

NER

TensorFlow

Relation Extraction

Computational Pathology

Histopathology Image Classification

Pinecone

PostgreSQL

CrewAI

Fine-tuning

Django

LangGraph

Docker

FastAPI

ChromaDB

MCP

AWS

n8n

JavaScript

Projects

MynEdge Maternal AI Platform

Lead Researcherprofessional

AI-driven maternal-health platform combining NLP, risk-prediction models, and conversational AI for low-resource clinical settings. Won Regional Winner at the Pakistan Ignite National Technology Fund Championship.

NLPRisk-prediction modelsConversational AI

GRAB-Rank

Researcheracademic

BERT + knowledge-graph-embedding framework for biomedical document ranking.

BERTKnowledge Graph Embeddings

OIE4KGC

Researcheracademic

Open information-extraction pipeline for automated knowledge-graph construction from unstructured biomedical text.

Open Information ExtractionKnowledge Graph Construction

Education

Master of Science (MS)

National University of Sciences and Technology (NUST)Software Engineering2015 – 2018

Thesis. Automated curation and ranking of biomedical literature using transformer models and knowledge graph embeddings

Coursework

Machine learningSentiment analysisSpam detection

Bachelor of Science (BS)

Ghulam Ishaq Khan Institute (GIKI)Software Engineering2010 – 2014

Recognition

Awards

Regional Winner, MynEdge maternal AI platform

Pakistan Ignite National Technology Fund Championship2025

Government-backed innovation grant

TMRP UK Doctoral Training Partnership Scholarship

TMRP UK2019

Competitive national doctoral funding

Google Summer of Code

Google2018

Open-source contribution under expert mentorship

NUST Masters Scholarship

National University of Sciences and Technology (NUST)

Full merit award

GCE A-Levels Scholarship

Beaconhouse School System

Patents & publications

Query Resolution of Literature Knowledge Graphs Using Hybrid Embeddings

publication2022

SGAI 2022, Lecture Notes in Computer Science, Springer.

With Coenen, F., Gamble, C., Kearney, A., Williamson, P.

GRAB-Rank: Document Ranking Using BERT and Knowledge Graph Embeddings

publication2021

DaWaK 2021, Lecture Notes in Computer Science, Springer.

With Bollegala, D., Coenen, F., Gamble, C., Kearney, A., Williamson, P.

OIE4KGC: Open Information Extraction for Knowledge Graph Construction

publication2020

DEXA 2020, Lecture Notes in Computer Science, Springer.

With Kearney, A., Gamble, C., Coenen, F., Williamson, P.

Maintaining Curated Document Databases Using a Learning to Rank Model

publication2020

SGAI 2020, Lecture Notes in Artificial Intelligence, Springer.

With Bollegala, D., Coenen, F., Gamble, C., Kearney, A., Williamson, P.

Temporal Spam Identification: A Multifaceted Approach to Identifying Review Spam

publication2018

IntelliSys 2018, Advances in Intelligent Systems and Computing Vol. 2, Springer. (94% accuracy; 223 reads)

With Qamar, U., Khan, F.H.

A Systematic Review of Time Series Based Spam Identification Techniques

publication2018

IntelliSys 2018, Advances in Intelligent Systems and Computing Vol. 1, Springer.

With Qamar, U., Noureen, R.

InstaSent: A Novel Framework for Sentiment Analysis Based on Instagram Selfies

publication2018

IntelliSys 2018, Advances in Intelligent Systems and Computing Vol. 1, Springer.

With Noureen, R., Qamar, U., Khan, F.H.

Social impact & activities

AI Bootcamp Lead

atomcampmentor2024 – Present

Co-led applied NLP/LLM/RAG bootcamp for 50+ professionals.

Course Instructor

Code First Girls UKmentor2021 – 2024

Delivered data science/ML courses to 85+ women entering tech.

Hire IqraReach out about a role, a contract or a conversation.For recruiters

Iqra's twin is AI, it can make mistakes.

More people on Kanvis

Browse more people Create your own Kanvis