Muhammad Ahmed A.

Key Skills

Software

Label Studio

Top Subject Matter

Medical imaging (3D/AR tool for mandibular tumor surgery, landmark detection)

Healthcare LLM chatbot / precision medicine knowledge retrieval

BioTech

Top Data Types

Image

Text

Document

Top Task Types

Fine-tuning

Tracking

Text Generation

Segmentation

Classification

Object Detection

Freelancer Overview

AI/ML Scientist. Currently working on a RAG architecture as part of an Ed-Tech startup - gaining experience in RAG/LLMs and deployment. Past Experience includes a computer vision research internship where I developed a screen-rendered augmented reality system. Adept at designing, training, testing, and deploying neural networks. Data annotation expertise using CVAT, 3D Slicer, and LabelStudio. Education includes Masters in Artificial Intelligence, University of Burgundy (2025) and Bachelor of Health Sciences, National University of Medical Sciences (2021).

Labeling Experience

Associate AI Division (Machine Learning, LLMs) – MedAngle (Remote, Pakistan)

Text Generation

Built retrieval-augmented generation (RAG) pipelines for a medical chatbot, which requires curating and using retrieved text contexts alongside model responses. Conducted alpha testing with end users and provided feedback to guide iterative improvements in the AI system. Ensured outputs were shaped by retrieval context to better support user queries in the healthcare domain. • Developed RAG pipelines using ChromaDB and LangChain • Integrated OpenAI for response generation • Provided feedback from alpha testing with ~150 users • Supported deployment considerations in resource-constrained environments

2025 - Present

Research Intern (Computer Vision, Medical Imaging) – ICMUB UMR CNRS 6302 Lab

Label StudioFine-tuningTracking

Performed manual annotation on a small-scale medical imaging dataset to support an AI workflow for landmark detection and registration. Used the annotated dataset as ground truth before auto-labelling with a YOLO model. Contributed labeled data to improve accuracy, including achieving target registration errors below 3 mm. • Manual annotation using Label Studio • Auto-labelling workflow using YOLO • Support for object tracking and image registration pipeline • Dataset preparation for semi-automated landmark detection

2025 - 2025

Clinical Data Analyst - National University of Medical Sciences

Label StudioImageSegmentationClassification

You analyzed healthcare data to support disease and treatment outcome prediction using Excel, MySQL, and Pandas. You improved radiographic lesion visualization and reduced treatment planning errors through image processing. You also applied machine learning to estimate one-year implant survival rates and used patient variables for risk stratification. • Performed data analysis and predictive modeling workflows • Enhanced lesion visualization and reduced planning errors by 20% • Conducted image processing and data annotation for CT and X-ray planning • Modeled implant survival outcomes to improve post-operative care planning

2022 - 2024

Education

Work History