Hi, I'm Mouhcine
Final-year AI and Machine Learning Engineering student at EMSI. Building production AI systems - deep learning models at 90%+ accuracy, RAG pipelines, and full-stack AI apps. Currently interning at Aishore as a Data and AI Engineer, building data-driven solutions powered by AI. Active on GitHub.
About
01I'm a final-year AI and Machine Learning Engineering student at EMSI in Casablanca. My focus is the gap between research and product: deep learning models that ship to production, RAG systems that actually retrieve the right context, and AI-powered apps that handle real users.
Recent proof: a cardiac disease classifier at 91% accuracy through feature engineering and PCA, a local RAG chatbot indexing 2,000+ documents at 92% retrieval precision, and MockCoder - an AI interview simulator now serving 200+ active users. Currently interning at Aishore as a Data and AI Engineer, building data-driven solutions powered by AI.
Work Experience
02Working on production AI and data systems, including LLM-based document extraction pipelines and structured-output validation.
Casablanca, MA
Highlights
- Built a 3-stage extraction pipeline (statement locator, text extraction, LLM structured output) that converts 200+ page financial-statement PDFs into schema-validated JSON in under 25s, parallelizing API calls across a 4-worker thread pool with queue-based live log streaming.
- Implemented context caching and a self-correcting validation loop that feeds Pydantic schema errors back to the LLM for corrected output, cutting input-token costs 75% and improving time-to-first-token.
Stack
Built a web platform to showcase Moroccan cultural heritage - end-to-end ownership from the Spring Boot API through the Next.js front-end and the AWS deployment.
Casablanca, MA
Highlights
- Designed and shipped the platform with Spring Boot, Next.js, PostgreSQL, and AWS S3 for media storage.
- Implemented a secure content-management system with role-based access for administrators.
- Built the admin authentication flow (JWT + refresh tokens) with protected routes on both the API and the front-end.
Stack
Automated invoice processing for the accounting team - removed manual data entry from the daily workflow using UiPath OCR.
Casablanca, MA
Highlights
- Cut manual data-entry time by 80% with a UiPath OCR pipeline that reads and validates incoming invoices.
- Integrated the bot end-to-end with the accounting system; team productivity went up ~50%.
- Owned the rollout: tested on a backlog of real invoices, handed off with runbooks and a monitoring dashboard.
Stack
Education
03Five-year engineering program in computer science with a focus on AI, cloud, and software engineering.
Casablanca, MA
Highlights
- Achievements: Deep learning, Software Engineering.
- Coursework: Artificial Intelligence, Cloud Computing, Data Structures, Databases.