Personal research project - Mostly vibe coded and now I'm refining it step by step | Python · Pandas · NumPy · Financial APIs · Time Series Analysis
A Python framework I built to explore quantitative approaches to financial market analysis. The focus is on systematic data extraction, time series modeling, and algorithmic pattern recognition across equities and crypto markets.
This is independent research — not a commercial product. Built to deepen my applied Python skills in a domain I find genuinely interesting.
Data Extraction
- Pulls historical OHLCV data from exchange APIs (Binance, KuCoin) and equity market feeds
- Structured pipeline from raw API response → clean DataFrame → analysis-ready dataset
Analysis Methods
- Time series analysis: trend detection, seasonality decomposition, anomaly flagging
- Price action and volume pattern recognition using Pandas and NumPy
- Fibonacci-based support/resistance level modeling
- Multi-asset correlation and segment analysis
Risk & Portfolio
- Systematic risk rules applied across asset classes
- Position sizing logic based on volatility metrics
- On-chain activity monitoring via blockchain explorers
Infrastructure
- Modular structure:
src/for core logic,scripts/for pipelines,tests/for validation - Dockerised environment for reproducibility
- Phase-based development with documented milestones in
docs/
| Layer | Tools |
|---|---|
| Language | Python 3.x |
| Data manipulation | Pandas, NumPy |
| Visualisation | Matplotlib |
| API clients | Binance API, KuCoin API |
| Environment | Jupyter Notebooks, Docker |
| Version control | Git |
├── src/ # Core analysis modules
├── scripts/ # Data extraction and pipeline scripts
├── tests/ # Unit and integration tests
├── docs/
│ └── roadmaps/
│ ├── strategic/ # Research planning and methodology
│ ├── phases/ # Phase completion documentation
│ └── technical/ # Algorithm implementation notes
├── config/ # Configuration files
├── reports/ # Output reports and results
├── requirements.txt
├── Dockerfile
└── docker-compose.yml
Active personal research project. Currently in beta — methodology is documented and functional, ongoing refinement of signal quality and test coverage.
Feedback and collaboration welcome. Open to discussion on methodology.
I have a formal background in database design and programming (SQL, Python, Java — IT Systems Management diploma) and spent 6 years doing data analysis and operations at Marley Spoon, where I worked on SaaS metrics, cohort analysis, churn modeling, and reporting infrastructure. This project extends that analytical work into financial time series.
Research enquiries: lavagante.project@gmail.com
Educational and research use only. Not financial advice.