Header
AI Democratization using ACUMOS AI Platform
ACUMOS AI Open Source Project offers the tools to create and deploy no-click AI. Check it out @ AI4EU Experiments!
Body
For the past few years, one of the things I have been building from the ground up is the Acumos AI Platform. Acumos is an Open-Source Linux Foundation project, which means that the software architecture and the code is available publicly for everyone in the world to use freely!
‘Making Artificial Intelligence Accessible To Everyone’ is the mission of the Acumos AI Platform Project. It is a platform and open-source framework that makes it easy to build, share, and deploy AI apps. Acumos standardizes the infrastructure stack and components required to run an out-of-the-box general AI environment. This frees Data scientists, AI modelers and Software Developers to focus on their core competencies and accelerates innovation.
Acumos has a lot of components and functionalities, and below is its architecture design — which we will not go into in this particular post.
And these were the technologies we used to build Acumos. All of them are open-source technologies.
So what exactly is Acumos and what does it offer?
Acumos is a multi-faceted platform that has many paradigms, that we will explore later in this series on Acumos. But at a high level -
-
Acumos is an AI marketplace i.e. a marketplace for AI, Machine Learning, LLM, NLP models like the Google PlayStore is for Android apps or the App Store is for iOS apps. In fact, Acumos is the world’s 1st open-source AI marketplace!
-
Acumos can help convert these AI/ML models to containerized Acumos microservice entities. This helps unlock many capabilities of the models and helps make the models easily consumable by business applications.
-
Using the Acumos Design Studio and the Blueprint Orchestrator, these microservice entities can also be composed with other heterogenous models i.e. models created in any language (Java, Python, R, etc.) or any framework (H2o.ai, Scikit learn, Spark, TensorFlow, R etc.) by anyone across the world to create complex AI pipelines that could perform a task better than the individual models or perform a task that the individual models cannot perform in isolation.
-
The Federation and Licensing features of Acumos can be used to discover, rate, review, share, publish, monetize, buy, and sell these AI/ML model entities between anyone and any organization across the world. For individuals, smaller orgs or NGOs who may not have the resources to hire ML engineers and data scientists and software professionals, it’s a way to crowdsource domain expertise and AI expertise.
-
These model entities as well as the Design studio created complex AI pipelines can be deployed to various cloud targets like Azure, AWS, GCP, Kubernetes from within Acumos with 1-click or be downloaded to local docker enabled machines as a set of communicating microservices.
-
The Acumos Platform itself can be simply and automatically installed for use locally using Acumos OneClick / All-in-One (AIO) or by only installing necessary components. Various different instances of Acumos belonging to Individuals, Universities, Companies (Vendors, Clients, or departments within the same company), Governments, NGOs, etc. anywhere in the world can be set up if required, to do Federation (sharing) and Licensing of assets contained in them. Thus, Acumos can conform to various Data Sovereignty requirements and Data Locality and Transfer restrictions like GDPR, CCPA, etc. by providing complete and fine-grained control over asset (model, data, document, metadata) sharing while also allowing the sharing of these assets and AI models between parties anywhere in the world!
-
Acumos has already been integrated with Open Network Automation Platform (ONAP), which is an already popular Open-Source Network Management Platform to seamlessly embed AI into the Network/Telecom/Security service design process itself, via DCAEMOD. I’ve had the great fortune of building both ONAP and DCAEMOD from scratch too, but more on them later!
So — coming back to Acumos, we have seen how it is one of the 1st Open-Source solutions that can be used to easily create and deploy no-click AI. Some of its software components themselves are novel and industry-firsts. I will discuss them in future posts.
Acumos was chosen and reused by the AI4EU project which is tasked with building the European AI On-Demand Ecosystem for their AI4EU Experiments Platform. It has empowered many AI creators, Data Scientists and Software Developers across the world Acumos is also commercialized by TechMahindra as the ‘GAiA Enterprise AI/ML platform.’ It has been featured on Forbes, LightReading Techcrunch and TelecomTV and other popular media.
Here are some illustrated papers discussing the usage and specific aspects of Acumos:
-
From Models to Microservices: Easily Operationalizing Machine Learning models
-
MLOps: Creating powerful AI pipelines by stitching together heterogeneous Machine Learning models
-
MLOps: Automatic, Zero-Touch and Reusable Machine Learning Training and Serving Pipelines
-
Easily Designing and Deploying AI enabled Network and Telecom services
Resources:
-
AI4EU Experiments Platform (Hosted by AI4EU, Create an account to access): https://aiexp.ai4europe.eu/#/home
-
Acumos Website : https://www.acumos.org/
-
Acumos Code: https://github.com/acumos
Author Bio:
Deven Panchal is an AI Leader currently with AT&T Labs Research, USA. He also serves as the Chief Innovation and IP/Patents Officer of the AT&T Innovation Network and is a Senior Member of the IEEE.
He has led the launch of 4 products into Open Source, being used worldwide today. He has delivered products and platforms projected to save ~1 billion$, enable Network Services creation, Network Management, Service Management and Orchestration and to Improve Customer service. He has helped build the world’s first AI/ML marketplace, was involved with AT&T’s first 5G VRAN/CRAN trial and delivered solutions that today power AT&T’s network as well as other telco networks. His work has benefited millions of people around the world, received multiple awards and has been featured on Forbes, TechCrunch, Light Reading, TelecomTV etc.
In the past, he has served as a Research Scientist at SAMEER, IIT Bombay building India's first indigenous LINAC and MLC for Cancer Radiotherapy.
Deven Panchal has been the Owner for several subprojects within Acumos, and has led global teams in the Design and Development of these components. He has written several papers on Acumos, and has contributed heavily to the official Acumos documentation. Additionally, he has been a Top Code Contributor and Peer reviewer for Acumos.
Feel free to reach out to him or use/cite his work.