Belaid Chaikhi

• Design, implement, and optimize machine learning models for classification, regression, and time series tasks, ensuring alignment with client needs and technical requirements. • Lead end-to-end development of machine learning, LLMs, and Retrieval-Augmented Generation (RAG) pipelines, including data extraction, preprocessing, model training, deployment, and monitoring. • Fine-tune LLMs using LoRA and other optimization methods to enhance model performance and adaptability to specific client needs. • Spearhead AI-driven automation solutions, including speech-to-text and audio classification projects, enhancing virtual employee capabilities for seamless communication. • Develop and optimize RAG models to extract similarities between entities across multiple domains, improving automated decision-making processes. • Lead NLP, OCR, and data extraction projects from unstructured data, enabling automated information retrieval and processing. • Conduct rigorous experimentation and performance evaluation using methods like cross-validation, hyperparameter tuning, and grid search to ensure optimal accuracy and efficiency. • Apply MLOps best practices, including version control with Git, Dockerized deployments, experiment tracking, and workflow optimization for scalable and reproducible results. • Collaborate with cross-functional teams, including software engineers, to integrate machine learning models into production systems via APIs, ensuring robust and high-performing solutions. • Stay up to date with advancements in machine learning, deep learning techniques, NLP, and speech processing, applying innovative methods to enhance AI model performance and business automation capabilities. • Leverage AWS cloud services for scalable infrastructure and ensure high availability of deployed AI solutions. • Tools: Python, PyTorch, TensorFlow, Scikit-learn, Numpy, Scipy, Flask, SQL, Pandas, Matplotlib, Plotly, LangChain, HuggingFace, Docker, MLFlow, AWS, Jira, NLP, RAG, Speech-to-Text, Whisper, Audio Classification, Time Series Analysis (ARIMA, Prophet, LSTM).

As a Machine Learning Engineer at Terra, my responsibilities include: Generating virtual images of waste. In-depth study of segmentation models. Improvement and optimization of existing models. Tools used: Python, OpenCV, NumPy, Matplotlib, Keras, TensorFlow, PyTorch.

As a machine learning engineer at Two-i, my role includes: State-of-the-art analysis of image semantic segmentation algorithms. State-of-the-art analysis dedicated to the training database. Implementation of the feasibility demonstration (POC) stage for a segmentation algorithm. Tools used: Python, C++, OpenCV, Keras, TensorFlow.