Before we start, I want to mention that this is a version of the same article I wrote in my personal blog. We are re-starting Data Intelligence consulting work in Latin America, and I thought it only proper to begin by talking about companies that inspire us and help us in our mission to bring Data Science and Business Analytics to companies — to help them bridge the gap between Europe, Asia, and the US.

I am a little jealous about a company I just discovered, called Nixtla. According to their own words, "Nixtla democratizes access to state-of-the-art predictive insights, eliminating the need for a dedicated team of machine learning engineers."

That is a tall order, and one I am a little doubtful about — because I am so used to people with minimal statistical skills basically butchering forecasting. But maybe that is just me and my luck in the real world.

The one thing I am truly envious of is their love for open source and their large codebase contribution to GitHub. For a company that wishes to democratize predictive analytics, they certainly put their money where their mouth is. The open way they are giving away code and examples is enough for more than one consultant to open a forecasting practice just on their excellent libraries.

Six Repositories of Predictive Goodness

I count six repositories with a plethora of predictive analytical code:

  • StatsForecast offers a collection of widely used univariate time series forecasting models, including automatic ARIMA, ETS, CES, and Theta modeling, optimized for high performance using Numba. It also includes a large battery of benchmarking models.
  • mlforecast is a framework for time series forecasting using machine learning models, with the option to scale to massive amounts of data using remote clusters.
  • NeuralForecast offers a large collection of neural forecasting models — from classic RNN and LSTM networks to the latest transformers: MLP, GRU, TCN, DeepAR, NBEATS, NBEATSx, NHITS, DLinear, NLinear, TFT, Informer, AutoFormer, FedFormer, PatchTST, StemGNN, and TimesNet.
  • HierarchicalForecast offers reconciliation methods including BottomUp, TopDown, MiddleOut, MinTrace, and ERM, with probabilistic coherent predictions including Normality, Bootstrap, and PERMBU.
  • tsfeatures calculates various features from time series data — a Python implementation of the R package of the same name.
  • Nixtla Open Source Time Series Ecosystem, a compendium of all the above, plus new classes.

And yes, I essentially listed the whole thing — not because I am lazy, but because I feel a little overtaken by this treasure chest of forecasting and predictive goodness. I wanted to give it the full space it deserves.

Technical Philanthropy in an Age of Proprietary Lock-In

In an age of more proprietary code and tools, such a degree of technical philanthropy is not unheard of, but it is not common either. The scale of Nixtla's contribution means that a consultant could build an entire forecasting practice on their libraries alone — without ever spending a dollar on commercial software.

I feel jealous of Nixtla for being a company so far advanced in predictive analytics that they can contribute to the wider world of data science and make a meaningful impact on the broader community. But then, that is certainly something worth being jealous of.

Follow Nixtla on their website and explore the repositories on GitHub.