Hasty Briefsbeta

Bilingual

Show HN: Python SDK – forecasting with foundation time-series and tabular models

4 months ago
  • #Time-Series Forecasting
  • #Tabular Inference
  • #Python SDK
  • FAIM SDK provides a production-ready Python SDK for time-series forecasting and tabular inference using foundation models.
  • Supports multiple foundation models including FlowState, Amazon Chronos 2.0, TiRex for time-series, and LimiX for tabular tasks.
  • Features include type-safe API with Pydantic validation, high performance with Apache Arrow serialization, and support for probabilistic & deterministic forecasts.
  • Includes async support for concurrent requests, rich error handling, and is production-ready with comprehensive error handling.
  • Provides detailed examples for time-series forecasting and tabular classification/regression tasks.
  • Supports batch processing, compression for large payloads, and connection pooling for optimal performance.
  • Available under Apache License 2.0 and includes citation information for research use.