Taiabur Rahman

Computer Vision & ML Engineer  |  Medical Imaging (AI)  |  Deep Learning  |  3D Image Analysis
📍 Paris 75019, France | ✉ taiaburbd@gmail.com | 📞 +33 07 66 93 69 20 | 🐙 github.com/taiaburbd | 🔗 linkedin.com
Profile — Computer Vision & ML Engineer with 1+ year in production AI/ML and 4+ years fullstack experience in healthcare SaaS. Erasmus Mundus Master in Medical Imaging (MAIA) — one of Europe's most selective AI programmes (acceptance rate <10%); research at INSERM Dijon on automated 3D brain segmentation (nnU-Net, PyTorch). Specialised in end-to-end deep learning pipelines for image segmentation, object detection and 3D volume analysis. Currently based in Paris — available immediately for on-site, hybrid or remote roles in France.
Professional Experience
Machine Learning Engineer Sept. 2025 – Present
BLACKBIRD.AI — New York, USA (Remote)
  • Fine-tuned LLM models (LLaMA, Mistral) on large-scale narrative intelligence datasets for misinformation detection
  • Built and maintained data pipelines processing 100K+ samples for model training and evaluation
  • Deployed and monitored models in production on AWS (EC2, S3) with FastAPI serving endpoints
  • Collaborated with research and engineering teams across time zones on model iteration and release cycles
Graduate Research Intern — Deep Learning & Neuroscience Feb. 2024 – Aug. 2024
INSERM UMR 1231, NEUROGEMM — Dijon, France
  • Designed an automated nnU-Net segmentation pipeline for high-resolution 3D histological mouse brain volumes, achieving strong Dice scores across multiple anatomical regions (Python, PyTorch)
  • Preprocessed and managed 400+ volumetric 3D imaging datasets on GPU cluster (SLURM); reduced manual preprocessing time by ~60%
  • Contributed to scientific documentation and presented results at internal INSERM research seminars
Full-Stack Developer (Full-time) Dec. 2017 – Aug. 2022
Vision Eye Hospital — Dhaka, Bangladesh
  • Designed and maintained a hospital ERP / HIMS platform serving 50+ clinical staff across prescriptions, diagnostics and surgical workflows
  • Reduced patient record retrieval time by ~40% through optimised database queries and caching
  • Managed production DevOps (Nginx/Apache, Linux) with 99%+ uptime over 4 years
Research Projects
3D Mouse Brain Segmentation — nnU-Net, PyTorch [github]
Automated segmentation pipeline for 400+ high-res histological 3D brain volumes; multi-region Dice scores competitive with manual annotation.
CT Image Registration — Elastix + SimpleITK [github]
Rigid and affine registration framework for multi-modal medical images; validated on public lung CT benchmarks.
3D Brain Tissue Segmentation (MRI) — TensorFlow, PyTorch, U-Net [github]
WM / GM / CSF delineation from volumetric MRI; achieved >90% Dice on test set using a custom 3D U-Net.
Skin Lesion Classification — VGG16, ResNet50, Keras [github]
Transfer learning on ISIC dermoscopic dataset; reached ~88% balanced accuracy with cross-validation fine-tuning.
Retinal Image Analysis (2D) — Classical CV pipeline [github]
Thresholding + morphological operations for hard exudate segmentation on DIARETDB1 dataset.
Education
Erasmus Mundus Joint Master
Medical Imaging & Applications (MAIA)
Sept. 2022 – Aug. 2024
Université de Bourgogne — Dijon, France
Università degli Studi di Cassino — Italy
Universitat de Girona — Spain
Medical image segmentation & registration, deep learning, computer-aided diagnosis, surgical robotics, e-health.
MSc — Software Engineering
Tianjin University, China
BSc — Computer Science & Engineering
Bangladesh University, Bangladesh
Technical Skills
ML / DLPyTorch, TensorFlow, Keras, Scikit-learn, nnU-Net
Computer VisionSegmentation, Registration, Classification, 3D Volume Analysis, OpenCV
AI AgentsLangChain, LangGraph, MCP, RAG, Tool Use, OpenAI / Gemini APIs
Medical Img.Segmentation, Registration, CAD, 3D volume analysis
Cloud / DataAWS (EC2, S3), GCP, Databricks
DevOpsDocker, CI/CD, Linux, Nginx
ToolsGit, REST API, Jupyter
Languages
Bengali — Native
English — Upper-intermediate (B2)
French — Elementary (A2, actively improving — B1 targeted by end 2025)
Training & Schools
17th EXCITE Summer School
Biomedical Imaging
ETH Zurich & Univ. of Zurich, Switzerland
Sept. 2023
Winter School 2023
Università di Cassino, Italy
Feb. 2023
Target Roles
  • Computer Vision Engineer (R&D / Production)
  • ML Engineer — Medical Imaging & HealthTech
  • Deep Learning Engineer — Image Analysis / 3D Vision
  • MLOps / AI Engineer (Docker, AWS, GCP, CI/CD)
ATS Keywords
Medical Imaging · 3D Segmentation · Deep Learning · Computer Vision · PyTorch · nnU-Net · SimpleITK · OpenCV · Python · FastAPI · Vue.js · React · Docker · AWS · GCP · Databricks · LLM · VLM · GPT-4 · Claude · Gemini · CLIP · LLaVA · Fine-tuning · Prompt Engineering · MCP · AI Agents · LangChain · LangGraph · RAG · CI/CD · MLOps · REST API · SaaS · INSERM · Erasmus Mundus
📍 Based in Paris, France — Available immediately for on-site / hybrid roles in France & remote roles worldwide