Provectus, an artificial intelligence (AI) company based in Silicon Valley, recently introduced Crystal Engine, a solution to the problem of predicting and personalizing customer behavior for start-ups, SMBs, and enterprises. The closed beta version of the product is now available.

Stepan Pushkarev, CEO of Provectus, said, “As a product, Crystal Engine falls somewhere between SaaS products already available on the market, and fully custom-built solutions. Finding the right balance between the versatility and flexibility of an AI solution is always a challenge. We strive to bring maximum value to our partners, so we created Crystal as a perfect fit to address two needs at the same time — minimizing time to market and meeting flexibility requirements for each particular business and product we work with. Our partners can benefit by gaining maximum accuracy of ML models while preserving ownership and security of their data and maintaining low TCO.”

Crystal Engine analyzes customer data to serve a variety of use-cases, such as:

  • Churn prediction
  • Customized recommendations
  • Next steps
  • For increasing conversion rates
  • Predicting Customer Engagement
  • Insights on the Customer Journey

Provectus believes Crystal Engine will be of particular value to subscription-based products across multiple categories and verticals, such as Financial Services and Insurance, Media and Entertainment, Healthcare, E-commerce and Retail, SaaS and Productivity Applications, and Marketplaces. Customer Retention and Customer Lifetime Value are the KPIs to improve using Crystal Engine.

Non-technical users can make customer predictions using Crystal Engine’s easy-to-use interface, free of coding requirements or installation requirements. Provectus emphasizes explainability in addition to simplicity. The results provide businesses with the insight they need to prove or disprove hypotheses, make product adjustments, and develop marketing strategies.

The product can be used for prediction purposes in a self-service mode (only for predictions), or customized and deployed on the client’s cloud for increased security and ownership.