Advanced Computing in the Age of AI | Wednesday, April 24, 2024

LogicPlum Upgrades its Enterprise AI Platform for Times-Series Predictions and More 

As barriers to entry are lowered, more companies are looking for ways to accelerate deployment and scaling of machine learning models. That is creating demand for MLOps tools and plug-play models used to accelerate data preparation and model creation.

To fill those needs enterprise AI platform vendor LogicPlum Inc. has unveiled an updated platform that automates development of time-series, or sequenced, models that among other new capabilities leverage trend analysis to predict the future value of a data series. LogicPlum said Monday (March 29) that the new LogicPlum 3.0 platform updates eliminate the need for costly machine learning infrastructure, data science and application development staffs.

The changes will allow for more application builders to develop and deploy predictive analytics and other time-series tools, according to the company. The updates include tools for sequencing, image visioning, MLOps and comparisons between information.

“We believe that giving a customer a machine learning model is not enough,” LogicPlum CTO Anoop Muraleedharan told EnterpriseAI. “They need a business outcome.”

To help customers achieve those goals, the three-year-old AI platform vendor focuses on “blended models” designed to reduce AI application development from days to minutes. In one example, Muraleedharan said LogicPlum’s platform was used by a healthcare provider to schedule and make specific shift assignments for nurses. The hospital supplied patient and staffing data which was used to create a detailed schedule for each shift.

LogicPlum's dashboard visualizes a range of data types.

Previously, complex hospital scheduling could take a month. LogicPlum claims its algorithm can make predictions on patient load, then assign and schedule nurses in a few minutes.

Along with time-series predictions, the updated platform also monitors model performance, accounting for “data drift” and the resulting need to retrain machine learning models to sustain accuracy.

Besides healthcare and staffing, LogicPlum’s customers also include retail and marketing firms. The platform’s predictive analytics are variously used for workflow applications such as predicting which patient is likely to cancel an appointment. The scheduling model would then be used to contact patients considered most likely to cancel.

The business outcome would be transferring the appointment slot to a waiting-list patient.

Along with predictive analytics, the platform is billed as plug-and-play so that it can be easily integrated into an existing workflow, eliminating the need for “beefy” infrastructure and application development teams. “All we need is the data,” said Muraleedharan.

The more data types that are applied to a model—ranging from text to images—the more accurate the model, according to the company.

LogicPlum is based in Franklin, Tennessee, and has offices in India and France.

About the author: George Leopold

George Leopold has written about science and technology for more than 30 years, focusing on electronics and aerospace technology. He previously served as executive editor of Electronic Engineering Times. Leopold is the author of "Calculated Risk: The Supersonic Life and Times of Gus Grissom" (Purdue University Press, 2016).