SaH Analytics International

• Led and developed multiple end‐to‐end ML/DL solutions (computer vision, NLP/GenAI, fraud detection), applying MLOps best practices.
• Satellite imagery analysis with computer vision: Designed, trained and optimized deep models for object detection, segmentation, change detection and land-cover classification from satellite images.
• Fraud detection: Partnered with Protegrity to build secure, privacy-preserving ML models for fraud detection.
• LLM development: Fine-tuned and deployed large language models using techniques such as instruction tuning, QLoRA, RLHF/PPO, DPO and ORPO.
• Agentic multimodal RAG pipeline: Implemented a retrieval-augmented generation system for text-based video interaction, leveraging multimodal embeddings, chunking, vector DB querying, re-ranking and function calling.
• APIs & UIs: Built robust REST APIs (FastAPI) and user interfaces (Gradio, Streamlit) for optimal UX.
• Workflow automation: Implemented CI/CD pipelines with GitHub Actions and automated ML workflows.
• Containerization & cloud deployment: Used Docker/Docker-Compose for scalable, resilient production deployments on GCP.
• Experiment tracking & monitoring: Adopted MLflow/ClearML for experiments and Prometheus/Grafana for production monitoring.
• Engineering best practices: Enforced coding standards, peer reviews and maintainability.
• GCP resource management: Administered cloud resources for training and deployment.
• Leadership & mentoring: Supervised ML engineer interns and delivered ML training sessions to clients and partners.
• Technology watch: Monitored advances in deep learning and generative AI.

Tools: PyTorch, Hugging Face, TorchGeo, transformers, YOLO, U-Net, ViT, Mask R-CNN, Detectron2, Ollama, Gemini, Llama, Mistral, Whisper, LangChain/Llamaindex, XGBoost, CatBoost, Random Forest, PyCaret, Poetry/UV, Git/GitHub, Docker, CI/CD, MLflow, ClearML, FastAPI, Streamlit, Gradio, Vertex AI, GCP, Open-webui.

Senior Machine Learning Engineer

Monk AI

Computer Vision R&D Engineer 

IFP Energies nouvelles

Machine Learning Research Engineer 

- Processed data for object detection tasks (Bounding box annotation, Putting annotations in the right format (YOLO, COCO or PASCAL VOC), anchor boxes generation using k-means).
- Benchmarked State-of-the-Art deep learning algorithms for object detection and segmentation.
- Developed a deep learning model for automatic image cropping.
- Developed a deep learning model for detecting small objects (Foraminifera) in geological images.
- Developed a deep learning model for instance segmenting rocks from images.
Tools : YOLO, RetinaNet, faster R-CNN, mask R-CNN, Cascade R-CNN, pytorch, detectron2, Microsoft VoTT, pytorchcv, keras, tensorflow, opencv, scikit-image

Deep Learning R&D Engineer - Computer Vision

Safran Aircraft Engines 

- Developed a probabilistic software for the quantification of uncertainties related to mechanical clearances in aircraft engines.
 - Deployed the software.
Tools : statistical modeling, regression, parametric and non-parametric density estimation, hypothesis testing, bootstrap, python, sklearn, numpy, pandas, Plotly Dash.

Internship in Probability & Statistics

Anacours

Private Tutor in Mathematics

Applied AI & ML Engineer with a PhD in Mathematics & CS and 5+ years of hands-on experience designing, deploying and scaling production-grade ML solutions (computer vision, NLP/GenAI, fraud detection). Proven track record partnering with cross-functional operations teams to translate complex datasets into predictive models and automated decision services. Proficiency in Python, MLOps (CI/CD, Docker, Docker-Compose, MLflow), financial-driven analytics, and clear communication of AI insights to non-technical stakeholders. Results-driven, client-focused, and adept at working in agile teams.

Taplio

Abdoulaye KOROKO's Linkedin Analytics

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Abdoulaye KOROKO's Linkedin Analytics

Want to drive more opportunities from LinkedIn?

Abdoulaye KOROKO's Linkedin Analytics