All projects

Team project · Python / Software Engineering

Time Series Task

An interface for time-series preparation, forecasting and result comparison.

  • Streamlit
  • pandas
  • FastAPI
  • Python
  • NumPy
  • statsmodels
  • Docker
Public repository

Project and contribution

Developed the data interface for a team time-series project: upload and editing, dataset preparation and charts. Integrated it with the forecasting API and refined the application structure.

Purpose

The application helps prepare time series, build forecasts and compare model results. Users upload and inspect datasets, configure parameters and explore forecasts and metrics through charts.

My work

  • Built Streamlit pages and organised them around user workflows.
  • Added data upload with delimiter selection and duplicate-row handling.
  • Refined dataset preparation and editing.
  • Connected the interface to the API and displayed analysis and forecast results.
  • Worked on charts and application structure.

Technical approach

Streamlit, pandas and NumPy support data preparation and the interface. The team solution includes a FastAPI service and forecasting methods such as SARIMA, ETS and Prophet, with manual and automatic parameter selection.

My contribution covered interface-side data work and server integration. This forecasting project is separate from SPAS System, which focuses on anomaly detection.