Mohamed ZWAWA

As the founder and lead Al engineer at VTAIK, I specialize in delivering cutting-edge Al solutions and Natural Language Processing (NLP) to empower businesses to harness the full potential of technology for innovation, efficiency, and digital transformation. addition overseeing the development of Al-driven applications, I provide professional training in Al and NLP, helping teams stay ahead with the latest advancements and tools.

Context : As an occasional instructor at Sup de Vinci, I conducted training sessions in Data Mining and Machine Learning for M1/M2 students in Big Data & AI. Combining pedagogy and technical expertise (Python, Scikit-learn, NLP, advanced modeling), I supervised modules covering exploratory analysis, dimensionality reduction (PCA, LDA), and machine learning algorithms, training students in key data science methodologies. Mission : • Teaching data mining and ML for M1/M2 Big Data & AI level. • Student supervision. Technical Environment : • Python, Scikit-learn, Pandas, Numpy, Matplotlib, Seaborn, Exploratory analysis • Decision modeling, Variable reduction, dimensionality reduction, PCA, LDA, Feature Selection • Machine Learning algorithms, RNN, CNN, Transformers, NLP

Context : As a key player in tracing the strategic roadmap, I collaborated with the business team to transform business needs into innovative Data solutions. Leveraging skills in artificial intelligence (LLM, fine-tuning, clustering) and advanced analysis (ML, semantics, prediction), I led strategic projects, combining innovation and impact on user experience. Mission : ● Squad Search Compréhension : Making DATA speak to improve user search experience. • Technical and functional scoping of Data Science projects • Development of a PoC ChatBot for PagesJaunes advice pages (GPT 4o / ELK / Embedding) • Development of a Search Agentic AI ChatBot (LangGraph, LangGraph Studio, Gemini Flash) • Semantic analysis of professional reviews (LLM / FineTuning SLM on Azure OpenAI) • API Management to secure Azure API access • Keyword clustering for search (ML) • Prediction of search trend evolutions (TRC) • Supporting business units on various Use Cases and gathering requirements • Participation in the annual Roadmap setup and mid-term review • Design and deployment of technical solutions • Design and maintenance of PagesJaunes knowledge base • Development of new user search interpretation • Migration to PagesJaunes' new search engine towards a Graph-oriented modeling (ArangoDB) and a Lucene engine • Participation in Agile Guilds & AI Guilds (presentations and monitoring) • R&D collaboration with the EXPRESSION team at IRISA • Supervision of a Data Scientist Squad SEO DATA & Partenariats : Increasing traffic on the portal https://www.pagesjaunes.fr • Rest API microservices for the portal (Java, Spring) • Rest API for partners (Bing, Yahoo, Apple, Amazon, Quant, Orange…) • Google Cloud Platform APIs • Python batch for backend Technical Environment : • GCP, BigQuery, Cloud Storage, Vertex AI, Gemini • JAVA, Spring, Spring Boot, Python, R, Spark, Hadoop, Pandas, Numpy, Scikitlearn • MongoDB, ElasticSearch, Apache Lucene, ArangoDB, MongoDB, Redis • Git, GitLab CICD, DVC, Airflow, MLflow, Kubernetes, Docker, Jenkins, Sonar • ML, KMeans, SVM, RNN, Transformers, Encoder-Decoder, PyTorch, Tensorflow • Azure OpenAI, LLM, SLM, Fine-tuning, Mistral.AI, Ollama, RAG, GenAI, GPT 4o, Embedding

Context : As a Tech Lead at La Poste, I led strategic projects aimed at modernizing the business tools for postal workers. Combining technical expertise (Java, PHP Symfony, API REST, OpenShift) and leadership, I ensured the technical and functional scoping and design of innovative platforms, such as a video broadcasting solution (TV Facteurs) and a digital library (E-Library), contributing to the company's digital transformation. Mission : ● Build of two La Poste Mobile projects : • TV Facteurs Project: Video broadcasting platform for postal workers. • E-Library Project: Platform for publishing documents (pdf, audio, or video) among postal workers. ● Technical and functional maturation. ● Design and technical scoping. ● Backend development. ● Operational maintenance. Technical Environment: • Java, PHP Symfony, API REST, Swagger • Mysql, Docker, GitLab, Checkmarks, OpenShift

Context : At Orange, I contributed to the stability and evolution of critical systems by managing incidents and ensuring the operational maintenance of key applications. My role focused on the evolutionary and corrective maintenance of strategic platforms (Chronos and ACTIV), combining backend expertise (Java, Spring, Oracle) and technical support to ensure reliable and efficient services. Mission : • N1/N2 incident management, resolution or escalation, on-site and remote support, on-call duty • Operational maintenance • TMA Configuration Tools – Chronos Project: Hosting services platform: Domain name, email, Fax… • TMA OSS – ACTIV Project: Application for managing technical incident files on the Orange network. Technical Environment : • JAVA, Spring, Struts, Bash, Perl • Oracle, MySQL, RabbitMQ, LDAP

Context : At Orange Labs, I participated in innovation around user experience by developing a personalized podcast discovery application. Leveraging skills in social mining and web development (HTML5, AngularJS, CakePHP), I designed a system using online activity data to tailor recommendations to user profiles, combining data science and development to enhance digital engagement. Mission : • Orange Radio Project: Design and development of a podcast discovery application based on user profiles. • Development of a podcast extractor. • Implementation of a user profile discovery system through their activities on social networks and the internet. Technical Environment : • Social mining, Weka, KNN • Podcasting, HTML5, AngularJS, CakePHP, MySQL..