How OneAdvanced Deployed Over 50 AI Agents on UK-Sovereign AWS

As a leading UK enterprise software provider, OneAdvanced has been at the forefront of innovation, leveraging cutting-edge technologies to drive business growth and improve customer experiences. In a recent project, the company successfully deployed over 50 AI agents on a UK-sovereign Amazon Web Services (AWS) platform. This achievement is a testament to OneAdvanced’s commitment to embracing artificial intelligence (AI) and its potential to transform industries. In this article, we will delve into the details of how OneAdvanced built a UK-sovereign AI platform, highlighting the key technologies and strategies employed in this groundbreaking project.

Introduction to OneAdvanced and its AI Initiative

OneAdvanced is a well-established enterprise software provider in the UK, serving a diverse range of industries with its innovative software solutions. The company has always been keen on exploring new technologies to enhance its products and services. The emergence of AI has presented OneAdvanced with exciting opportunities to revolutionize its offerings and improve customer engagement. By harnessing the power of AI, the company aims to develop more intelligent, automated, and personalized solutions that cater to the evolving needs of its clients.

Building a UK-Sovereign AI Platform

The decision to build a UK-sovereign AI platform was driven by the need to ensure data security, compliance, and sovereignty. OneAdvanced chose to self-host Llama 4 Maverick and Llama Guard 4 on Amazon SageMaker AI, a fully managed service that provides a robust environment for building, training, and deploying machine learning models. This approach allowed the company to maintain full control over its data and AI models, adhering to stringent UK data protection regulations.

Amazon SageMaker AI provides a comprehensive set of tools and services that enable data scientists and developers to build, train, and deploy AI models quickly and efficiently. By leveraging SageMaker AI, OneAdvanced was able to streamline its AI development process, reducing the time and effort required to build and deploy its AI models.

RAG Pipeline on pgvector

To further enhance its AI capabilities, OneAdvanced implemented a RAG (Retrieval, Augmentation, Generation) pipeline using pgvector. RAG is a powerful framework for building AI models that can retrieve relevant information, augment existing data, and generate new content. The integration of pgvector, a vector database, enabled OneAdvanced to efficiently store, search, and manage large volumes of vector data, which is essential for building high-performance AI models.

The RAG pipeline on pgvector allowed OneAdvanced to develop more sophisticated AI models that can retrieve relevant information from vast datasets, augment existing data with new insights, and generate high-quality content. This capability has opened up new possibilities for the company, enabling it to develop more intelligent and automated solutions that can drive business growth and improve customer satisfaction.

Strands Agents SDK on Amazon ECS

To build over 50 AI agents, OneAdvanced utilized the Strands Agents SDK on Amazon Elastic Container Service (ECS). The Strands Agents SDK provides a comprehensive set of tools and APIs that enable developers to build, deploy, and manage AI agents quickly and efficiently. Amazon ECS, a highly scalable and secure container orchestration service, allowed OneAdvanced to deploy its AI agents in a flexible and efficient manner.

The combination of Strands Agents SDK and Amazon ECS enabled OneAdvanced to build and deploy a large fleet of AI agents, each designed to perform specific tasks and functions. These AI agents can be easily scaled up or down to meet changing business requirements, ensuring that the company’s AI platform remains flexible and responsive to evolving customer needs.

Benefits and Outcomes

The successful deployment of over 50 AI agents on a UK-sovereign AWS platform has yielded numerous benefits for OneAdvanced. The company has been able to develop more intelligent and automated solutions that can drive business growth, improve customer satisfaction, and enhance overall competitiveness.

Some of the key outcomes of this project include:

• Improved Data Security and Sovereignty: By self-hosting its AI models on Amazon SageMaker AI, OneAdvanced has ensured that its data remains secure and sovereign, adhering to stringent UK data protection regulations.
• Enhanced AI Capabilities: The implementation of a RAG pipeline on pgvector has enabled OneAdvanced to develop more sophisticated AI models that can retrieve relevant information, augment existing data, and generate high-quality content.
• Increased Efficiency and Agility: The use of Strands Agents SDK on Amazon ECS has allowed OneAdvanced to build and deploy AI agents quickly and efficiently, reducing the time and effort required to develop and deploy new AI solutions.
• Scalability and Flexibility: The company’s AI platform can be easily scaled up or down to meet changing business requirements, ensuring that OneAdvanced remains responsive to evolving customer needs.

Conclusion

The successful deployment of over 50 AI agents on a UK-sovereign AWS platform is a significant achievement for OneAdvanced, demonstrating the company’s commitment to innovation and its ability to harness the power of AI to drive business growth. By leveraging cutting-edge technologies such as Amazon SageMaker AI, pgvector, and Strands Agents SDK, OneAdvanced has developed a robust and scalable AI platform that can support a wide range of applications and use cases.

As the adoption of AI continues to grow, OneAdvanced is well-positioned to remain at the forefront of innovation, leveraging its expertise and experience to develop more intelligent, automated, and personalized solutions that cater to the evolving needs of its clients. The company’s UK-sovereign AI platform will play a critical role in driving business growth, improving customer satisfaction, and enhancing overall competitiveness in the years to come.