Projects
Cloud ML Demand Forecasting Pipeline
FeaturedPythonAWS SageMakerXGBoostLambdaAPI GatewayS3boto3
Production ML pipeline with a layered S3 data architecture, XGBoost training on SageMaker, and a serverless inference API via Lambda and API Gateway. Simulates a multi-store retail demand forecasting system with seasonality and promotions.
R²: 0.84MAPE: 9.1%Real-time endpointServerless REST API
AI-Powered Knowledge Base Assistant
FeaturedNext.jsFastAPIPythonRedisDockerVector Search
AI-powered knowledge assistant using Next.js and FastAPI, enabling semantic retrieval over 200+ unstructured documents with vector search, Redis caching, and Dockerized services.
~45% latency reduction~30% efficiency boost200+ indexed documents
Image Preprocessing Techniques
PythonOpenCVNumPytkintermatplotlib
Image processing application with a tkinter GUI implementing histogram equalization and Gaussian convolution. Allows applying visual color styles and comparing results in real time.
Histogram equalizationGaussian filterInteractive GUI3 color styles