Data Science & AI Engineer (End-of-Studies Project)

Remax Wise

2026-02 - 2026-08

Led the PropIQ AI project, an assistant that verifies real‑estate compliance using the CRISP‑DM methodology from business analysis to deployment. Designed and implemented a RAG pipeline to index legal documents, create embeddings, and perform semantic search, significantly reducing hallucinations. Built a benchmark of seven LLMs tracked with MLflow to select the optimal model. Implemented MLOps with modular architecture, CI/CD pipelines via GitHub Actions, Docker containers, and monitoring dashboards. Developed a FastAPI backend with Firebase authentication and a Flutter mobile app, integrating a Quality‑of‑Life module for property rating.

Data Science & AI Intern

Mouvmnt

2025-06 - 2025-08

Created a legal chatbot for lawyers by building a RAG pipeline that grounds answers in verified legal sources, thereby minimizing hallucinations. Implemented filtering and targeted prompting to keep the chatbot focused on legal topics. Analyzed user query patterns to improve search relevance and answer quality. Developed backend APIs with FastAPI and containerized the solution with Docker, hosting code on GitHub.

Agentic AI Platform for Multi-Domain Data Analytics

2026-10 -

Designed and built a platform that lets users ask natural‑language questions about any dataset, converting them to SQL via a LangGraph AI agent and visualizing results with interactive Plotly dashboards. Implemented an MCP server offering eight tools for dataset inspection, data quality checks, SQL execution, and dashboard generation, all orchestrated through a FastAPI backend and Streamlit UI. Deployed LLMs with Ollama and ensured robustness with smoke tests and health‑check endpoints. Added support for large datasets by implementing on‑the‑fly chunking during upload, all containerized with Docker.