your123 - stock.adobe.com
Aiming at an audience of business and data analysts who don't want to do a lot of coding, predictive analytics startup Pecan AI has added one-click model deployment and live monitoring to its platform.
The vendor, which is based in Tel Aviv, Israel, and introduced the new features on Wednesday, develops a predictive analytics platform that automates the creation of predictive models that businesses can use to focus on customer journey KPIs and track end-user success with products, customer loyalty, and the gap between what customers expect from a product and the reality.
Pecan and competitors
While Pecan AI touts its automated machine learning technology as optimized for customer experience sales and supply chain applications, the vendor's targeted approach is not unique, said Forrester Research analyst Boris Evelson.
Other business intelligence platforms such as Microsoft Power BI and Salesforce already have tools non-data scientists can use for point-and-click, drag-and-drop analytics applications, Evelson noted.
However, Pecan's positioning of its platform fits into the market niche of AI analytics tools that would likely interest business intelligence analysts who are not data scientists with highly specialized skills.
Boris EvelsonAnalyst, Forrester Research
"If it's set up, pre-configured, prepackaged plug and play, then that absolutely that would be attractive,” Evelson said. "But again, everyone does that."
And in addition to providing a platform that can be set up for any use case, analytics vendors typically package at least a few preset programs for certain use cases, he noted.
Pecan's competitors include established bigger vendors such as SAP, DataRobot, Dataiku and Alteryx and more recent startups including Imply Data, Noogata and Unsupervised.
Although other tools on the market do similar things, Pecan has found a degree of success in the past few years. The vendor has raised $116 million for its predictive analytics platform in six rounds of fundraising since 2018, according to Crunchbase.
Technologies like Pecan's are also popular now due to the movement to "democratize" AI by making it more accessible to business users. However, these tools do not eliminate the need for professional data scientists, Evelson said.
For example, although Pecan says its software can measure accuracy drift [comparing predicted values to actual values over time], "a nonprofessional is not going to know what's good, what's bad in accuracy drift," Evelson added. "This whole notion of democratizing AI and machine learning to nonprofessionals is still very early."
Since many organizations find it hard to hire data scientists, the Pecan platform closes the gap between business and data scientists by enabling business and data analysts to easily create predictive AI models, said cofounder and CTO Noam Brezis.
"The most suitable people in the organization are [data] analysts to create AI models," Brezis said.