Rayane Kherzi

• Design and deployment of an interactive web application for managing and searching electoral contacts, with a Node.js/Express backend and SQLite database, hosted on Fly.io for scalability and robustness.

• Advanced exploitation of public electoral data to create predictive models, using machine learning techniques (regression, decision trees) to anticipate local demographic and electoral trends.

• Automation of analyses and visualizations to enable precise targeting of campaign messages, optimizing activist actions based on data-driven insights.

• Complete project documentation and technical recommendations to secure data, improve persistence, and plan for production deployment with robust solutions.

• Participation in digital transformation projects for large corporate clients, digital transformation, consulting, data consulting, process digitalization, large accounts, project management, banking sector, energy sector, public sector.

• Development, deployment, and monitoring of artificial intelligence models in production environments, MLOps, DevOps, CI/CD, Azure ML, AWS SageMaker, model monitoring, DataOps, Docker containerization, versioning, scalability, ML workflow automation.

• Integration and development of AI APIs for business applications and conversational chatbots (RAG, LLM, NLP), NLP, LLM, REST APIs, FastAPI, Flask, microservices orchestration, RAG chatbot, conversational AI, GCP, Azure API Management, API security, microservices, Swagger documentation.

• Advanced analysis and visualization of customer data, datalakes, Power BI, Tableau, ETL, SQL, data profiling, data visualization, dynamic reporting, data mining, KPIs, business dashboards, predictive analysis.

• Management of continuous integration (CI/CD) and automated deployment of AI solutions, Jenkins, GitLab CI, Azure DevOps, CI/CD pipelines, automation, monitoring, unit tests, Release Management, continuous deployment, cloud infrastructure.

Alignment of entities using RREA applied to knowledge graphs - Cergy, entity alignment, RREA, knowledge graph, graph embedding, KG, machine learning, data science, academic research, AI research, graph neural network.

Using a GAT (Graph Attention Network) to align entities between heterogeneous graphs, using triplets (entity, relation, entity), GAT, Graph Attention Network, entity matching, triplet modeling, PyTorch, TensorFlow, deep learning, graph data, graph representation, graph matching, relational data, embedding model.

Implementing a semi-supervised learning approach, with iterative integration of aligned entity pairs, semi-supervised learning, weak supervision, iterative training, data augmentation, model refinement, Python, ML pipeline, algorithm design, active learning, workflow automation.

Performance evaluation via CSLS; adjustment with Hits@1/5/10 and MRR metrics, CSLS, Hits@K, MRR, metric evaluation, model validation, benchmarking, performance metrics, competitive analysis, PyTorch Lightning, research publication, reporting, hyperparameter tuning.