Structured Classifications

Feature Generation

Chain model predictions and improve your classifications.

As with any software module, Machine Learning models can be pipelined. The output of model A can become the input for model B. For instance, think of an Outbound Automation model, that detects whether a company might be interesting for your sales team. An input for this model, that can significantly improve the prediction quality is the industry of a company. This is a model chain. The output of the industry model becomes a feature, i.e. an input, for the outbound automation.

Of course, as such intelligent feature generators require plenty of training data, it becomes quite difficult to quickly build and test such ideas. This is where refinery enters the game. Our system is designed to support you in your data-centric ideas by quickly scaling your data labeling, so that you can focus on building pipelines. Also, our system helps managing your data. You can slice and monitor your data on any kind of feature combination, data enrichment or heuristic, helping you to find potential weak spots in your data.

If you’re interested, you can try out our open-source version. We’re more than happy to help you along the way, supporting you with your use case.

Become a data pioneer now

We are building tools for the age of data-centric AI.
Let's build great use cases together.