# Nixtla — AI-Powered Time Series Forecasting & Anomaly Detection
> https://www.nixtla.io

Nixtla is an enterprise-grade time series intelligence platform. We provide TimeGPT, the first foundation model for time series, enabling zero-shot forecasting, anomaly detection, and demand prediction without training data. Trusted by Microsoft, Lyft, Prudential, Decathlon, and leading enterprises worldwide.

## Products

### TimeGPT
The first foundation model for time series forecasting. Zero-shot predictions, anomaly detection, and demand forecasting via a simple API. No training data required.
- Zero-shot forecasting (no training needed)
- Anomaly detection at scale
- Multi-horizon predictions with conformal intervals
- Exogenous variables support
- Fine-tuning on custom data
- SOC 2 Type II compliant
https://www.nixtla.io/

### TimeGPT-2
Next generation foundation model with improved accuracy and expanded capabilities for time series analysis.
https://www.nixtla.io/timegpt-2-waitlist

### Open Source Ecosystem (49M+ downloads)
- **StatsForecast**: Lightning-fast statistical forecasting (AutoARIMA, ETS, CES, MSTL, Theta)
- **NeuralForecast**: Deep learning models (NBEATS, NHITS, PatchTST, TimesNet, TFT)
- **MLForecast**: Machine learning forecasting (LightGBM, XGBoost, scikit-learn)
- **HierarchicalForecast**: Hierarchical reconciliation methods
- **CoreForecast**: High-performance C/C++ backend
- **UtilsForecast**: Common utilities for time series processing
https://github.com/Nixtla

## Deployment Options
- **Cloud API**: Managed service with instant access
- **Self-hosted**: Deploy on your own infrastructure (AWS, Azure, GCP)
- **On-premises**: Air-gapped deployments for regulated industries

## Industries Served
- **Supply Chain**: Demand forecasting at SKU/location level, inventory optimization
- **Energy**: Electricity load, renewable generation, and price forecasting
- **Financial Markets**: Risk management, portfolio optimization, market analysis
- **Retail**: Sales forecasting, promotion planning, seasonal demand
- **Healthcare**: Patient volume, resource allocation forecasting

## Key Metrics & Results
- 85% reduction in false alerts (Lyft case study)
- 10x improvement in anomaly detection speed
- 500+ ML models onboarded at scale
- 50% higher accuracy vs traditional methods
- 95% codebase reduction in forecasting pipelines
- IDC Market Leader for 3 consecutive years

## Documentation
Complete guides, API references, and tutorials for implementing time series forecasting.
https://www.nixtla.io/docs

## Blog
Technical articles, research updates, and practical guides on forecasting, anomaly detection, and predictive analytics.
https://www.nixtla.io/blog

## Success Stories
Real-world case studies from Microsoft, Lyft, Prudential, Decathlon, and other enterprises.
https://www.nixtla.io/success-stories

## Company
Founded in 2021. Backed by M12 (Microsoft's Venture Fund), True Ventures, GPV, and Energize. 16k+ GitHub stars, 49M+ downloads, 20k+ TimeGPT users, 10k+ community members.
https://www.nixtla.io/company

## Contact
- Support: support@nixtla.io
- Sales & Enterprise: https://www.nixtla.io/book-a-call
- GitHub: https://github.com/Nixtla
- X/Twitter: https://x.com/nixtlainc
- LinkedIn: https://www.linkedin.com/company/nixtlainc

## Optional
- llms-full.txt: https://www.nixtla.io/llms-full.txt