AT&T logo AT&T Labs Research

AT&T’s Open Source collaborations are driving innovation, making it easier than ever to deploy technology into the marketplace.

Network AI

Open source software allows us to manage more cost-effectively, better control and more quickly deploy network services than ever before. We’ve announced several major open source initiatives, all designed to maximize speed and innovation while controlling costs, including ONAP, DANOS, Acumos and Akraino Edge Stack.

Through these initiatives, we’re delivering functioning software, tools, applications and platforms. Our engineers and developers are contributing code that’s being used in AT&T and around the world.

With the number of open source projects at AT&T continuing to expand, we are building a framework for how they integrate with and support one another.

Think of this as Network AI, an intelligent software-defined framework for these projects. AT&T Labs will spearhead this initiative, with a focus on identifying areas where a combination of software, open source and AT&T resources can drive innovation for the industry.

Acumos

AT&T wants to make artificial intelligence (AI) accessible to everyone. To that end, we’re collaborating with Tech Mahindra to build an open source AI platform called Acumos, which makes it easy to build, share and deploy AI applications. Hosted by the Linux Foundation, the platform will provide a marketplace for accessing, using and enhancing those applications.

The Acumos Property Assistant is a machine learning model that allows users to price their home based on fundamental property characteristics. Chris Buonocore, developer of the Acumos Property Assistant, was awarded $50,000 as winner of the Acumos AI Challenge, sponsored by AT&T and Tech Mahindra.

Resources and News

Download Acumos White Paper »

Open Source and Edge Technology

Exploring the future of open source technology. Watch as experts discuss what is next for open source technology and how it will transform the AT&T network.

Exploring the future of edge technology. Listen as experts discuss how users should think about edge computing and what it means for the future.

Akraino

Akraino is a new open source project intended to create an open source software stack supporting high-availability cloud services optimized for edge computing systems and applications. The project will offer users new levels of flexibility to scale edge cloud services quickly, to maximize the applications or subscribers supported on each server and to help ensure the reliability of systems that must be up at all times.

Resources and News


Nanocubes

Nanocubes allows us to interact with large data sets through visualization. Three key innovations have made this possible: software breakthroughs at AT&T Labs, the power of the Cloud and exposed APIs. Nanocubes can be applied to many diverse datasets, which is why AT&T released it into open source for others to build upon and use.

RCloud

A collaborative platform built in the cloud that enables developers to see and deploy code their peers have already created across coding languages.

Additional Open Source Projects

See more open source projects here.

Technical Publications

Packaging and Sharing Machine Learning Models via the Acumos Open Platform
Shuai Zhao, Manoop Talasila, Guy Jacobson, Cristian Borcea, Syed Anwar Aftab, John Murray, Mazin Gilbert
International Conference on Machine Learning and Applications, 2018

Control Loop Automation Management Platform (CLAMP)
Mazin Gilbert; Rittwik Jana; Eric Noel; Vijay Gopalakrishnan
IEEE Global Conference on Information Processing, 2016

Autonomous Services Composition in Domain 2
Rittwik Jana; Mazin Gilbert; Syed anwar Aftab; Farheen Cefalu; Pamela Dragosh; Serban Jora; Thomas Kirk; John Lucas; Arthur Martella; John Murray; Sundar Ramalingam; Christopher Rath; Shu Shi; Richard Wright; Avi Zahavi
IEEE Global Conference on Information Processing, 2016