New! We're expanding PwCs' agent OS to AWS, bringing Bedrock and AWS-native tools to the enterprise. I'm thrilled to announce that we are bringing agent OS to AWS, with integration for AWS-native services for document intelligence, storage and orchestration, as well as agent monitoring. PwC’s agent OS is an enterprise AI command center, seamlessly connecting and scaling intelligent agents into business-ready workflows, up to 10x faster than traditional methods. We're seeing tremendous success with PwC clients, many of whom on standardizing their agent deployments on agent OS, so it's great to expand that audience further with these new capabilities: 🪣 S3: Allows your agents access to secure file storage and retrieval within S3 buckets (including the new S3 Vectors). Great for getting context, or setting memory. 🔍 OpenSearch: Power high-performance search across structured and unstructured data, supporting full-text queries, filtering and aggregations. Perfect for RAG. 💬 Textract: Uses ML to extract structured data (forms, tables) from PDFs and scanned documents, supporting pre-processing steps in intelligent workflows, and additional agents context. 🤘 Bedrock: Connects agent OS agents to foundation models like Claude, Mistral, Llama, and Nova, for document drafting, translation, summarization, reasoning, math, research, and more. ⚙️ MCP servers: Makes AWS-connected tools callable from within custom workflows using the MCP protocol — enabling seamless tool calling for agents. This dramatically expands the capabilities of all agents with your existing fleet of AWS services and tools. With this launch, we’re making it even easier for enterprises to deploy powerful AI agents inside the cloud environments they already trust—securely, at scale, and with real impact. Excited to see what our clients build next. Onwards!
Fantastic!!
Bringing agent OS to AWS unlocks a powerful bridge between enterprise-grade AI and proven cloud infrastructure. This could accelerate real-world adoption across industries.
love this.
I am surprised Textract is part of the stack here. There is a lot of spacial context that is lost when you just have the text from the page, and templated post processing strategies don't scale for most enterprise documents use cases.
That was fast, love the integrations!
Chief Commercial Officer, Mayfield | Empowering Entrepreneurs to Scale Successful Ventures | Accelerating Product-Market Fit and Early Customer Adoption | Connecting CIOs, CTOs, and CXOs to Drive Corporate Innovation
23hExcellent work Matt