Tupl and Premo announce joint presentation at DES 2022
2022-06-03 · Press Releases
Tupl, a market leader in automating wireless network operations, today launched the Tupl ML Toolkit as part of the TuplOS AI Engine to provide wireless network engineers with the ability to build machine learning models seamlessly. Engineers simply route their data into a machine learning model, which learns from engineers’ examples and automates their complex investigation tasks. Machine learning models can automate tasks to streamline network operations and improve customer experience. During a beta period, customers tested the Tupl ML Toolkit functionality and were able to generate a machine learning model that delivered results within hours of deployment. Starting today, the Tupl ML Toolkit module of TuplOS AI Engine is available globally to wireless companies through Tupl Operation Automation use cases.
Machine learning models are very challenging to implement given that the builder requires both domain knowledge of the relevant data to be consumed by the model and a solid understanding of the techniques that help the model effectively learn from that data. In wireless operations, engineers would greatly benefit from machine learning models that automate and scale highly complex tasks using their data, yet engineers typically aren’t experts in model design. The Tupl ML Toolkit enables the creation of machine learning models by engineers with little to no experience in this field. Additionally, application integrators who want to extend application functionality would also vastly benefit from the Tupl ML Toolkit.
Petri Hautakangas, Tupl CEO, said: “The Tupl ML Toolkit is our way of bringing our machine learning technologies to engineers in an easy-to-use solution. Our Toolkit helps improve the speed and consistency of engineering decisions, and facilitates the identification of previously unforeseen issues. Additionally, the Tupl ML Toolkit has a wide array of applications that include use cases for IoT and automated management of other complex networks, such as utilities and enterprise. We look forward to seeing how our customers leverage the toolkit to streamline their operations.”
Pablo Tapia, Tupl CTO, said: “What’s unique about the methodology in our solutions is the simplicity of digitalizing engineering knowledge by a combination of supervised and unsupervised learning techniques. With our Toolkit we enable customers to keep full visibility on their decisions, which provides a continuous cycle of model refinement that moves away from a ‘black box’ approach. The Tupl ML Toolkit gives engineers that transparency, and it also facilitates the overall data cleaning and preparation process to turn raw data into machine learning features, a process that typically takes 60-80% of the overall effort in a machine learning exercise. By offering this toolkit, we hope to enable the growth and adoption of more machine learning in the wireless industry.”
Tupl offers AI and machine learning-based solutions that dramatically reduce operational costs through efficiencies in network issue detection and resolution. Tupl’s AI Engine, TuplOS, turns the telecom operator’s engineering experience into a digital knowledge base that is leveraged to build use-cases that can then be the basis to automate and improve existing processes.
To learn more about the Tupl ML Toolkit, visit www.tupl.com.
Founded in 2014 by telecom, big data and AI veterans, Tupl is transforming customer access and experience in the telecom industry thorough improving operations with leading wireless operators across the US, Canada, Japan, Mexico, and Europe. Its AI Engine, TuplOS, utilizes machine learning and several other utilities to enable faster innovation cycles for network and customer care operations. Tupl is headquartered in the US in Bellevue, Washington with presence in Spain, Mexico and Japan, and is continuing its rapid global expansion in 2018. To learn more about Tupl and request a demo, visit www.tupl.com.
2022-06-03 · Press Releases
2022-05-13 · Press Releases
2022-05-05 · Press Releases