// CS Engineering Student
I build intelligent systems that automate the complex, from ML pipelines and RAG architectures to fullstack AI products, so humans can focus on what truly matters.
# about.md
I'm Hakim, a Computer Science engineering student passionate about the intersection of AI and software engineering. I specialize in building intelligent, automated systems, from machine learning pipelines to fullstack AI powered products.
My work spans AI/ML model development, fullstack web applications, and automation workflows using modern tools like LangChain, RAG architectures, FastAPI, and React.
I love turning complex problems into elegant, production ready solutions.
$ ls -la ~/projects
Automates SEO content creation and social media publishing end-to-end. Analyzes competitor blogs, generates optimized articles with metadata, and schedules posts across platforms.
HR automation system for candidate screening and interview scheduling. Vectorizes CVs, evaluates job fit via RAG, assists recruiters with an AI agent, and automates calendar booking.
Web app for event organizers to auto-generate and send hundreds of personalized certificates. Features an intuitive editor and an AI module that generates certificate designs and content automatically.
Python FastAPI app that automates public sector tender evaluation through an 8 phase pipeline combining deterministic logic with LLM powered document analysis. Uses the "Enforcer + Reasoner + Decider" pattern.
Fullstack ML app detecting fraudulent transactions with Logistic Regression, undersampling for imbalanced data, containerized with Docker Compose, and comprehensive Pytest coverage.
Content based filtering using TF IDF and Cosine Similarity across genres, keywords, cast, and director. Dynamic in-memory similarity matrix with a dark glassmorphism UI.
Logistic Regression + TF IDF vectorization for spam classification. Premium dark mode glassmorphism UI, modular backend, structured logging, and production Docker infrastructure.
Draw on HTML5 canvas or upload an image to classify handwritten digits. TensorFlow/Keras MNIST model auto-trained during Docker build. Zero local dependencies required.
RandomForestRegressor achieving ~98.9% R² accuracy on Gold ETF (GLD) price forecasting from market indicators. Pydantic validation, structured logging, and fully Dockerized deployment.
SVM classifier trained on the PIMA Diabetes Dataset achieving ~77% test accuracy. Sleek dark mode UI, modular ML pipeline with separate training and inference engines.
RandomForest model with dynamic probability gauge, handles class imbalance via imbalanced learn, interactive predictor for tweaking demographics, and production grade artifact caching.
XGBoost Regression achieving ~0.83 R² on California housing data. Geographic and demographic processing, ASGI lifespan model loading, CORS security with Pydantic, Nginx frontend.
$ echo "Let's build something."
Have a project in mind or want to collaborate? I'd love to hear from you.
