“They were getting the data, but were unable to understand driver behavior and forecast demand. Where the chargers are to be installed, and how many. Such insights were needed to maximize revenue. They went live within a month. Every week, they could go and talk to their customers and say this is how your employees are using the chargers. You have less chargers at this location. Formcept helped ChargePoint sell more.”
In the battery monitoring industry, traditionally, they show the electrical values but what would the user do with that? We built a product called Battery Doctor with Formcept. It consumes all the electrical values. Uses AI and ML to process it. And provides recommendations such as: ‘Within the next 4 to 5 months, you need to replace this particular battery in this location”. And here, I am talking of 20 to 30,000 batteries.
As a procurement intelligence company, we deal with a lot of data from multiple different sources and formats. We needed a layer on top of our database that is intelligent enough to understand the data relationships, ‘X data is related to Y, Y it is related to Z.’ That is what enables us to give alerts to users so they make their best next decision.
There are more than 30 pre-packaged plugins (such as clustering, recommendation, correlation, classification, pattern mining, regression, extractor, and transformer) and more can be added as and when needed.
Formcept automatically creates the ML pipeline, based on the preparatory transformations and ML algorithms you apply on your training set. You just need to review the outputs and the quality at each step.
Formcept offers visual comparisons as well as contextual summaries depending on the algorithm chosen. You do not need to work with JSON or write codes to visualize the model performance.
Formcept documentation hosts comprehensive information on the parameters to evaluate and the comparable algorithms for each model. See our documentation.
Yes, Formcept has a sticky AI agent that serves as a contextual assistant across your ML workflow. You can use it to get suggestions and to evaluate a model.
Yes, Formcept allows you to automatically run ML algorithms periodically. And lets you set up alerts so you are notified when a process is completed or stuck. You can also set up reports on model performance so you can spot and fix quality issues at the earliest.
Formcept uses Apache Spark, which makes it horizontally scalable. This means, whether you work with 1Mn records or 1Bn records, you can train and run any kind of ML models on your data with the same speed and smoothness.
Yes, Formcept allows you to bring your ML pipelines from other tools. Alternatively, you can even export the ML pipelines you built in Formcept to other tools.
