UK-Based Company Deploys Over 50 AI Agents on Sovereign AWS Infrastructure
In brief
- OneAdvanced, a UK-based enterprise software company serving over 10,000 customers in industries like healthcare and legal services, has successfully deployed more than 50 AI agents on Amazon Web Services (AWS) infrastructure within the United Kingdom.
- This move ensures that all data remains within UK borders, meeting strict compliance requirements for sectors handling sensitive information.
- The deployment uses open-source models Llama 4 Maverick and Llama Guard 4, self-hosted on AWS SageMaker, alongside a Retrieval Augmented Generation pipeline and specialized agents built with Strands Agents SDK.
- The decision was driven by the need to maintain data sovereignty and adhere to stringent regulatory standards.
- OneAdvanced’s customers in regulated industries require clear data residency and security assurances.
- While initial prototyping used Amazon Bedrock, the company switched to self-hosting models to meet these requirements since desired models weren’t available in the UK region at the time.
- This approach involved building a production-grade solution with features like content moderation and document retrieval.
- This deployment highlights how businesses can achieve compliance while leveraging AI without relying on third-party services.
- OneAdvanced’s architecture integrates Amazon Aurora PostgreSQL with pgvector extension and runs agents on Amazon ECS, ensuring scalability and performance.
- As more companies seek to comply with data residency laws, such solutions may pave the way for similar deployments globally.
Terms in this brief
- SageMaker
- A service by AWS that provides tools for building, deploying, and managing machine learning models. It's used here to host AI models locally.
- Retrieval Augmented Generation
- An approach where AI systems combine external data retrieval with generation tasks, improving their ability to answer questions based on context.
- Strands Agents SDK
- A software development kit for building intelligent agents that can perform tasks like information retrieval and decision-making. It's used here to create specialized AI agents.
- pgvector
- An extension for PostgreSQL that allows efficient vector similarity searches, useful for tasks like recommendation systems or document retrieval.
Read full story at AWS ML Blog →
More briefs
SpaceX Acquires AI Coding Startup Cursor
SpaceX has acquired AI coding startup Cursor. The acquisition started in April when the companies teamed up for model training efforts. Cursor will now have access to a large fleet of GPUs to build and train better models at a lower cost. SpaceX paid $60 billion for the purchase. The companies have released two models together, Grok 4.5 and Grok 4.6, which can do tasks like coding and computer-aided design. The acquisition will help Cursor develop low-cost AI models with the help of SpaceX's computing power. SpaceX will continue to develop new models with Cursor.
USA Today Partners with Palantir
USA Today partnered with AI firm Palantir to monetize audience data. The deal was revealed during an earnings call and caught over 800 unionized journalists off guard. Journalists object to the deal because of Palantir's contract with Immigration and Customs Enforcement. They say this creates a conflict of interest and raises data security concerns. The company wants to use reader data for subscriptions and advertising. The deal has sparked a heated debate about data monetization and transparency. The future of the partnership is uncertain and will likely be watched closely by the media industry.
AI Investments Shift as Nvidia Reduces Bet on OpenAI, Anthropic Defies Bubble Fears
Nvidia has significantly reduced its financial guarantee for OpenAI's Ohio data center project, cutting it from $250 billion to just under $120 billion. This move comes after investors expressed concerns about the risks involved. Meanwhile, Anthropic is challenging the notion of an AI "bubble" with impressive revenue growth-jumping from $4.7 billion to $11.5 billion in a single quarter. The shift in investor sentiment toward OpenAI reflects broader caution in the tech sector. While some worry about overvaluation and potential risks, Anthropic's strong financial performance suggests that at least one AI company is thriving despite these concerns. This divergence highlights the varying fortunes within the AI industry, where some players are flourishing while others face pressure to reassess their strategies. As the AI landscape continues to evolve, keeping an eye on how other companies adapt to market conditions and investor demands will be crucial. The interplay between financial caution and technological progress is likely to shape the future of the AI sector in unexpected ways.
Banks Rely on Few AI Providers
Banks are using AI to automate tasks and help with operations. More than 75% of City companies now use AI. They use it to process insurance claims and assess customers' creditworthiness. This matters because it creates a systemic dependency. A problem at one provider could spread quickly across customers and sectors. Regulators may increase their focus on operational resilience and third-party concentration in the AI model stack. The AI race also risks creating vendor dependence risk, meaning a few providers could control the price of AI services. Banks will try to offset these risks by negotiating contracts and using open-source AI models. Banks will continue to invest in AI to cut costs and increase revenues.
Israeli Startup Corma Raises $60 Million to Fight AI Cyberattacks
Corma has raised $60 million to build AI designed to fight AI-powered cyberattacks. The company is developing a foundation model for defensive cybersecurity. The company is working with Fortune 100 and Fortune 500 companies. Its systems reduced threat-response times by more than 94% in early deployments. Corma's AI agents operate across an organization's existing security tools and processes. Corma will use the funding to continue developing its AI models for defensive cybersecurity. The company will keep working to protect organizations from cyberattacks.