Spotify Batch Data Processing
Python project for extracting and processing data from Spotify Web API with OAuth 2.0 authentication and pagination strategies.
Deep Learning · Data Engineering · AWS · LLM · RAG
PyTorch, AWS Glue, Airflow, LangChain, RAG LinksAll projects
Overview
Tagantino AI is a smart agriculture platform that helps Moroccan farmers better manage water resources by forecasting droughts using satellite data and providing real-time decision support through an LLM assistant. Built between January and June 2025.
Key Components
ETL Pipeline
- Data Sources: CHIRPS (Climate Hazards Group InfraRed Precipitation with Station data)
- Processing: AWS Glue ETL for data transformation and cleaning
- Orchestration: Apache Airflow for workflow management
Deep Learning Models
- Drought forecasting models trained on historical satellite data
- Time-series prediction for precipitation and soil moisture levels
LLM Assistant
- RAG Architecture: Retrieval-Augmented Generation for contextual responses
- LangChain Integration: For building conversational AI workflows
- Real-time decision support for irrigation and crop planning
Technologies
- AI/ML: Python, PyTorch, Hugging Face Transformers
- Data Engineering: AWS Glue ETL, Apache Airflow
- LLM: LangChain, RAG
- Frontend: Next.js, TypeScript
Impact
Helps farmers in water-scarce regions make data-driven decisions about irrigation, potentially reducing water waste and improving crop yields.