Decompose your demand data at scale for FREE

Securely share your anonymized demand data and we’ll automatically decompose your data by detecting weekly and monthly cyclicities, seasonal patterns, long term and short term trends, and remainders. No risk. No commitment.

Or, feel free to simply send us a note to learn more.

Doing it on your own takes weeks. PredictHQ will do it for you in seconds, for free.


Save time

We use an automated machine learning model to automatically decompose your demand data into baseline and residuals. The best part? It can run at scale, which means we can process thousands of time series in seconds.

Immediately use the output

After your demand data is analyzed, you will have full visibility into your baseline demand and anomalies. You will get 4 key outputs: the date, actual demand, baseline demand, and the remainder.


Unlike typical decomposition methods, our decomposition engine has been built with an energy-based nonparametric ML model. This means you aren't limited by certain hypotheses restricting your data.

Checklist of what you'll need

  • Daily aggregated demand time series data with > 6 month duration
  • < 20% missing rate in each demand time series data set
  • No more than 7 consecutive days with missing demand values
  • Start date of demand data to be later than 01-01-2017
  • Positive demand values only
  • Non-constant demand time series

Demand decomposition can be complex and take time. Let us do the work for you and in a fraction of the time.

We know that teams can do this on their own, but it can often take up to mutltiple weeks to break out all of the time series.

We have created this automated decomposition tool to be able to quickly understand demand, patterns, remainders, and trends at scale.

Simple. Accurate. Free.

Request your FREE Decomposition