zeineb romthana

• Designed and introduced the team’s first reusable ReAct agent architecture using LangGraph and LangChain, defining core patterns for prompts, tool orchestration, memory flow, and graph structure that later became the foundation for all internal AI agents. • Built 2 production RAG pipelines (DevOps, FinOps) using Qdrant + bge embeddings + reranking, reaching 88%+ retrieval precision to eliminate dependency on paid search APIs. • Engineered a FinOps anomaly detection system (ETL from FOCUS/cloud APIs, PyOD ensemble: IForest + HBOS + LOF) and a Prophet-based spending forecast, giving clients early visibility into budget overruns and anomalous spend. • Designed the local L

Skills: English

Develop a chatbot to assist with financial inquiries with the GenAI team at BCG

As a Data Scientist Intern, I worked on developing an AI-powered voice chatbot for a sports application designed to provide users with personalized fitness coaching and support. The primary goal was to create an engaging and interactive experience that enhances users' fitness journeys through real-time feedback and personalized recommendations. 1)Responsibilities: - Data Collection and Preparation:*Gathered and structured datasets including Q&A for exercise instructions, motivational phrases, and progress tracking queries. - NLP and Speech Recognition:Utilized Natural Language Processing (NLP) techniques for intent recognition and entity extraction, and integrated speech recognition and text-to-speech (TTS) systems to facilitate voice interactions. - Model Training and Evaluation:Employed machine learning models to handle user queries, provide accurate exercise guidance, and deliver motivational responses. 3)Tools and Techniques: -Models: GEMMA,mistral7b,llama3 -Pgraming languages :Python - NLP Libraries:NLTK, spaCy - Speech Recognition 4) Goals: -Voice Chatbot Development: Accuracy: 90%+ in understanding exercise instructions andResponse Time:Under 2 seconds -User Engagement:Interactions: 500+ unique daily interactions in 3 months and Motivational Messages: 10,000+ messages in the first quarter. -Progress Tracking: 95% of relevant queries answered.

analyse the gas sales and dashboard creation

Task 1: Web scraping to gain company insights Task 2: Predicting customer buying behaviour

-Increased employee productivity by 50% by automating attendance recording with a facial recognition application. -Improved security measures by 30% through the implementation of facial recognition technology.

Skills: Lead Qualification · Microsoft Excel · Oral Communication · Customer Relationship Management (CRM) · Presentations · Software as a Service (SaaS) · Solution-oriented · Problem Solving · Relationship Development · Research Skills · Excel Dashboards · Auditing · Teamwork

The Entity Control Board (ECB) is in charge of ensuring compliance to the governance system of AIESEC, as well as managing the quality cases raised against entities and violations to the AIESEC Exchange Program Policies and the AIESEC Portfolio Internal Policies.

I was honored to work with Doro Packaging Company on my final-year project with my partners; I had such a wonderful experience, from hard skills to soft skills.

improving the work flow using Siemens Tia portal

i had an outstanding experience with IEEE organization in the field of robotics and i gain a lot of expreience and a remarkable improvement

Skills: Siemens Tia Portal · Oral Communication · Software as a Service (SaaS) · Automation · Observation · Teamwork