Pecan AI can now automate the deployment and monitoring of predictive AI models
Predictive analytics startup Pecan AI Ltd. said today it’s bringing one-click model deployment and integration to common customer relationship platforms, marketing automation tools and other business systems.
By doing so, it promises to make life much easier for business teams to take action on accurate predictions about future churn, demand and other customer conversion metrics its platform generates.
Israel-based Pecan AI sells a data science platform that enables companies to build artificial intelligence models for tasks such as predicting customer demand and finding opportunities to reduce operational costs. Historically, building a custom neural network required specialized technical know-how, as well as a considerable amount of time and effort. Pecan AI says that its platform allows companies to deploy AI models in weeks without the need for specialized machine learning expertise.
Pecan AI’s models are able to generate individualized predictions for every customer and send this information to whatever business system a company is using, be it Salesforce, Dynamics 365 or something else. In this way, companies can carry out more precise actions such as serving ads to specific customers based on their history.
The company said the automatic deployment of AI models is possible due to its automated label engineering capabilities. With most other platforms, data scientists have to create, monitor and update data labels on an ongoing basis.
It’s a time-consuming, manual process that Pecan AI eliminates by creating labels based on a specific use case. By acting as the label engineer, Pecan AI said it can provide deeper insights into what’s happening within a business and therefore generate more accurate predictions.
A second useful feature of Pecan AI is its ability to model the performance of these AI models in real time. The platform continuously monitors live models for signs of degradation, the company explained. AI models have a tendency to degrade over time due to internal changes in consumer behavior, plus external changes in data integrity. As such, AI models need to be monitored to ensure they’re still predicting things accurately.
Pecan AI said it can automate this process to ensure its models retain their high level of predictive accuracy without the need for specialist data engineers. Should there be a need to change course, due to model drift, data leakage or something else, the platform can even provide recommendations on what must be done. The new capabilities allow Pecan AI to automate every aspect of AI model creation and deployment, including data cleansing, designing and building the models, deploying them, and then ensuring continuous data input and output to client systems.
Holger Mueller of Constellation Research Inc. told SiliconANGLE that it’s hard enough to deploy AI models at the best of times, and even more so when it involves deploying them inside business applications.
“It’s good to see Pecan AI simplifying this process with the famous one-click deployment model,” Mueller said. “We’ll have to see the feedback from early deployments to be sure, but the value proposition sounds very tempting.”
Pecan AI co-founder and Chief Technology Officer Noam Brezis said data science models traditionally take many months to build, train and test. In addition, it can take many months to prepare and connect the data needed to train those models, he said.
“By automating the process and ensuring models are easy to deploy and monitor for value delivery against business goals, Pecan’s latest platform enhancements are helping customers not just build and successfully deploy more production-grade models, but ensure they yield better quality predictions,” he said.
Image: Pecan AI
A message from John Furrier, co-founder of SiliconANGLE:
Your vote of support is important to us and it helps us keep the content FREE.
One click below supports our mission to provide free, deep, and relevant content.
Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.