Iguazio

Quadient, a leading provider of omnichannel customer experience solutions, needed a way to unify and combine every single data type they work with to create machine learning applications that run in real time. This would enable its developers to build a SAAS cloud platform for enterprises to improve customer experience and achieve digital transformation through business automation. Its platform would then be able to predict events by ingesting data from several sources including real-time sensor data, historical data (like ERP and CRM), news, social media, flight tracking, and other sources — and then leverage AI and ML to interact with all that layered data to derive business interaction predictions. This enabled Quadient to provide new capabilities and additional value to its clients (particularly those in the insurance space). Prior to finding and leveraging Iguazio, Quadient tested several cloud platforms to unify, store, and provide a single interface for the various data types it wanted to work with (such as relational data, key values, time series, etc.)

Home Credit International (HCI), a global consumer financial provider, recognized the potential of Machine Learning (ML) models in financial institutions, particularly in risk-related use cases. However, they faced challenges in deploying ML models efficiently. The time to delivery was long and access to data was limited. HCI’s internal research revealed that nearly 80% of the time spent on data science-related tasks was dedicated to collecting datasets and cleaning and organizing the data, leaving only about 20% of the time for core tasks like building training sets, mining data, and refining algorithms. In 2021, the average delivery time of an AI initiative, from prototype to production, was more than seven months. The biggest blocker for more efficient use of AI/ML was access to data, followed by the need for a proper AI/ML environment.

Ecolab, a global leader in water, hygiene, and infection prevention solutions, wanted to develop predictive risk models for water systems, industrial machinery, and other applications. The company's machine learning journey began in 2016 with a project to develop bacterial growth risk models using existing sensor data. However, the process of building, deploying, and maintaining machine learning models in production was complex and challenging. The company needed a data science collaboration platform that would bring together its large, geographically dispersed team, while efficiently using cloud computing resources. The deployment of machine learning models at Ecolab followed a 'rewrite-and-deploy' pattern, where model development occurred independent of the application developers. This approach led to deployment timelines exceeding 12 months on average.

Payoneer, a digital payment platform, needed a way to serve its AI/ML fraud predictive/preventative models against fresh, real-time data to provide their customers with a safer payment experience. The company was using a retroactive approach that detected fraud attempts after the fact, which meant customers could only block users after a (possibly successful) fraud attempt. This approach had several limitations including the inability to track fraud attempts across complex networks, lack of advanced analytics and log correlation to identify anomalies, and a negative impact on customer experience and satisfaction. Payoneer needed a solution that leveraged sophisticated algorithms to track multiple parameters and detect fraud within complex networks. While Payoneer had built sophisticated machine learning models, these only worked offline and could not be used against real-time threats.

NetApp, a leading provider of hybrid cloud data services, needed to enhance its Active IQ solution to incorporate an AI-driven digital advisor. The goal was to use AI to gain intelligent insights into its customers’ storage controllers and deliver prescriptive guidance, as well as automate “best actions” to achieve predictive maintenance on said devices. The company was dealing with the challenge of analyzing 10 trillion data points per month from storage sensors worldwide. The existing infrastructure of Active IQ, built on Hadoop, was unable to cost-effectively enable real-time predictive AI, run large-scale analytics, or deploy new AI services at scale. The traditional data warehouse and Hadoop-based data lake were unable to efficiently process the trillions of data points collected from storage controllers at the speed required to derive actionable intelligence needed for real-time predictive maintenance.