Amazon Bedrock for AI Professionals and Developers

We've been diving deep into Amazon Bedrock over the past couple of months, exploring the fascinating capabilities it unlocks for our customers. 💭 Some of you may remember my skepticism when the preview was announced in April of this year… Well, I’m happy to report that I was wrong. After getting hands-on-keyboard with the service over the last two months, I now firmly believe this service propels AWS ahead of the curve and paves the way for the democratization of GenAI. Bedrock gives builders unfettered access to LLMs from multiple providers through a consistent API that’s deeply integrated with AWS. 🚀     🌟 My favorite features?   ✅ Speed. We’ve been getting about 20 tokens per second on Claude V2, not accounting for network latency. On Claude Instant, we’ve seen 100s of tokens per second. ✅ Scale. Despite taxing the service pretty aggressively, we have yet to hit any rate limits. ✅ Growing set of model options. So far at Caylent, we’ve been working with Antrhopic’s Claude V2, Stability’s Stable Diffusion XL v2, and AI21’s Jurassic-2. Last Wednesday, AWS announced the addition of Cohere’s Command model, which I can’t wait to try. ✅ Privacy. Your data is never used to retrain the models for other customers. No inference request’s input or output is used to train any model. Model deployments are inside an AWS account owned and operated by the Bedrock service team. Model vendors have no access to customer data. ✅ Security. You can customize the FMs privately and retain control over how your data is used and encrypted. Your data, prompts, and responses are all encrypted in transit (TLS 1.2) and at rest with AES-256 KMS keys. You can use PrivateLink to connect Bedrock to your VPCs. Your data never leaves the region you’re using Bedrock in. IAM integration enables RBAC, ABAC, and resource-based policies that allow your organization to customize access based on your organizational policies. ✅AWS Integration. For existing AWS customers, the deep integration of Bedrock into tooling like CloudWatch, CloudTrail, and IAM means Bedrock is production ready as soon as it’s generally available.   💼 We’ve given 100+ demos of Bedrock over the last 60 days, and it’s thrilling to see customers start to move beyond experimentation and into production. All of these demos and customer conversations led to the creation of our Generative AI Knowledge Base Catalyst that connects Amazon Bedrock with Amazon Kendra to deliver bespoke enterprise scale retrieval augmented generation capabilities to any AWS Customer. This is already powering our internal knowledge base at Caylent and even providing weekly summaries of updates. 🔜 What's next on the horizon? I'm eagerly awaiting access to Bedrock's game-changing feature, Agents.   💡With all the above, it's no wonder we're thrilled to help customers #MoveToBedrock and #BuildOnBedrock.   #GenAI #AWS #AWSBedrock

Every AI agent needs the same scaffolding underneath: tools to interact with, an environment to work in, a system to manage context, memory to help personalize, identity to control access, and observability to understand and course correct. Wrapped in a loop that calls the model, picks a tool, and recovers from failures. This is the agent harness, and until now, every team built it from scratch. Today we launched the 𝗺𝗮𝗻𝗮𝗴𝗲𝗱 𝗮𝗴𝗲𝗻𝘁 𝗵𝗮𝗿𝗻𝗲𝘀𝘀 𝗶𝗻 𝗔𝗺𝗮𝘇𝗼𝗻 𝗕𝗲𝗱𝗿𝗼𝗰𝗸 𝗔𝗴𝗲𝗻𝘁𝗖𝗼𝗿𝗲 (public preview). You declare your agent, and run it in three API calls. You point to the model, tools, skills, and instructions as configurations in the API, and AgentCore stitches together everything around it to make the agent production-ready. What you get out of the box: 1️⃣ 𝗔𝗻𝘆 𝗺𝗼𝗱𝗲𝗹, 𝘀𝘄𝗶𝘁𝗰𝗵 𝗺𝗶𝗱-𝘀𝗲𝘀𝘀𝗶𝗼𝗻. Bedrock, Anthropic, OpenAI, Gemini, or any OpenAI-compatible endpoint (coming soon). Switch providers mid-session without losing context. 2️⃣ 𝗧𝗼𝗼𝗹𝘀, 𝗱𝗲𝗰𝗹𝗮𝗿𝗮𝘁𝗶𝘃𝗲𝗹𝘆. MCP servers, AgentCore Gateway, built-in Browser and Code Interpreter, or your own inline functions. One config line per tool - no boilerplate code to write. 3️⃣ 𝗦𝘁𝗮𝘁𝗲𝗳𝘂𝗹 𝗯𝘆 𝗱𝗲𝗳𝗮𝘂𝗹𝘁. Each session runs in a secure, isolated microVM with its own filesystem and shell. Short-term and long-term memory persist across sessions. 4️⃣ 𝗜𝗻𝘁𝗲𝗿𝗮𝗰𝘁 𝘄𝗶𝘁𝗵 𝘁𝗵𝗲 𝗲𝗻𝘃𝗶𝗿𝗼𝗻𝗺𝗲𝗻𝘁 𝗱𝗶𝗿𝗲𝗰𝘁𝗹𝘆. Run shell commands on the session's dedicated microVM to set up repos, extract artifacts, or debug. 5️⃣ 𝗕𝗿𝗶𝗻𝗴 𝘆𝗼𝘂𝗿 𝗼𝘄𝗻 𝗰𝗼𝗻𝘁𝗮𝗶𝗻𝗲𝗿. Pre-bake source code, runtimes, and dependencies. The harness wraps your environment and works with it. 6️⃣ 𝗕𝗿𝗶𝗻𝗴 𝘆𝗼𝘂𝗿 𝗼𝘄𝗻 𝗦𝗸𝗶𝗹𝗹𝘀. Compose your agent with Agent Skills: bundles of markdown and scripts that give it domain knowledge on demand. Use the open ecosystem or write your own. The harness handles loading and execution. 7️⃣ 𝗕𝘂𝗶𝗹𝘁 𝗼𝗻 𝗼𝗽𝗲𝗻 𝘀𝗼𝘂𝗿𝗰𝗲. The harness is powered by Strands Agents, AWS's open-source framework. When config stops being enough, export to code and keep running on the same compute, same microVM, same observability. No re-architecture, no platform tax. Trying a new model or tool is a config change, not a code rewrite. Managing context, remembering across users, enforcing policies, using a new skill: again config, not infrastructure. Weeks of plumbing collapse into minutes! Learn more in our docs: https://lnkd.in/gqv5NmW3 and in our GitHub samples: https://lnkd.in/gKWysZkD #aws #bedrock #agentcore #harness

I've been using 𝗔𝗺𝗮𝘇𝗼𝗻 𝗕𝗲𝗱𝗿𝗼𝗰𝗸 𝗔𝗴𝗲𝗻𝘁𝗖𝗼𝗿𝗲 in almost every customer project since it launched in October 2025. It's become my go-to for taking AI agents from prototype to production without drowning in infrastructure work. #AgentCore gives me #serverless agent #hosting, managed memory, tool integration, auth, observability, and policy enforcement. All as composable services. I pick what I need, plug it in, and move on to the actual problem. But when I talk about AgentCore to customers and colleagues, I notice the same thing every time. People get lost. Nine services, two categories, a lot of capabilities. It's a lot to absorb in a conversation. I figured this is probably the case for most people working in the AI agent space right now. So I created a visual cheat sheet to make it easier. 𝗜𝗻 𝘁𝗵𝗶𝘀 𝗱𝗶𝗮𝗴𝗿𝗮𝗺 𝗜 𝗰𝗼𝘃𝗲𝗿:  • All 9 AgentCore services grouped by what they actually do (Build vs Assess)  • The architecture flow with Runtime as the hub  • A quick reference table: "I need to X then use Y" One page. Everything you need to understand AgentCore and decide which services fit your use case. #AmazonBedrock #AIAgents

GenAI Architecture – Week 7 Project 7: Enterprise RAG for messy, real-world documents When I first started experimenting with RAG, everything looked easy — clean PDFs, perfect datasets, and sample code that “just worked.” But when I tried applying it in the enterprise world, reality hit hard. I remember sitting with a stack of documents: scanned contracts with faded text, multi-column PDFs where tables broke alignment, and 100+ page policy docs that felt like they were written in a different century. My “neat” RAG pipeline just collapsed. That’s when it clicked: enterprise RAG isn’t about retrieval — it’s about resilience. So in Week 7 of my 10-week GenAI Architecture series, I’m sharing how I built an approach to handle messy, real-world enterprise data. How the architecture works" - A developer ( poking around in Kiro or Cursor IDE) fires off a query. - Retriever Agent uses hybrid embeddings in vector DBs (Chroma, Redis, FAISS) to widen the net. - Enterprise Data Layer swallows PDFs, DOCs, scanned files, intranet pages — basically, the jungle of enterprise knowledge. - GroundX + EyeLevel step in to make sense of chaos: layouts, tables, multi-column pages, even OCR corrections. - Amazon Bedrock AgentCore, with Claude/Nova, ties it all together — interpreting, reasoning, and giving back a response that feels business-ready. Why this matters Because real data is messy. And unless your RAG system is built to handle it, the answers you get will be brittle. This design gives - Confidence that no doc was “too ugly” for the pipeline - A way to explain why an answer came back the way it did - The flexibility to run this at both small team scale and enterprise scale Where I see this shine - Legal/compliance teams are swimming in decades of contracts - Internal policy Q&A across multiple, outdated intranet sites - Research copilots digging through regulatory docs 🛠 Tech Stack Kiro IDE | Cursor IDE | GroundX | EyeLevel | AWS Bedrock AgentCore | Claude | Nova | ChromaDB | Redis | FAISS | SQLite | S3 📚 For more tips and deeper dives, check out my books: - Generative AI for Software Developers: https://lnkd.in/gy8vU-Nr - AWS for Solutions Architect (3rd Edition): https://lnkd.in/gsR_qMEN Looking back, this was the week when I stopped thinking of RAG as a “cool demo” and started seeing it as enterprise infrastructure. Next up → Week 8: Federated data with MindsDB + MCP Unified Server. ⚡ #GenAI #AWSBedrock #AgentCore #Claude #Nova #EnterpriseAI #RAG #GroundX #EyeLevel #VectorDB #10WeeksOfGenAI #KiroIDE #CursorIDE

Agentic AI's landscape is evolving so quickly! These intelligent, autonomous agents can perceive, reason, and act independently to achieve complex goals. AWS Prescriptive Guidance (July 2025) provides a roadmap for organizations to implement them effectively and securely. Key Highlights • Frameworks • Strands Agents: Model-first design, MCP integration, native AWS service support • LangChain and LangGraph: Graph-based workflows, multimodal processing, rich orchestration • CrewAI: Role-based, multi-agent orchestration mirroring human teams • Amazon Bedrock Agents: Fully managed, with action groups and built-in observability • AutoGen: Conversational, asynchronous, human-in-the-loop and code execution • Protocols • Model Context Protocol (MCP): Open standard for interoperability and OAuth security • A2A (Google) and AutoGen (Microsoft): Alternatives, with MCP recommended for production • Tools • Protocol-based: MCP SDKs (Python, TypeScript, Java) • Framework-native: Strands, LangChain, LlamaIndex • Meta-tools: Workflow, memory, and agent graph for advanced orchestration Who Should Take Note • Cloud architects building scalable AI workflows • Developers and ML teams integrating Bedrock, OpenAI, or Anthropic Claude • Enterprise leaders deciding between managed and DIY frameworks • Compliance officers ensuring secure and interoperable AI adoption Noteworthy Aspects • AWS positions MCP as the backbone for open, secure agent communication • Strands Agents powers real-world modernization (AWS Transform for .NET) • CrewAI with Bedrock demonstrates up to 90 percent faster enterprise automation flows • LangGraph and AutoGen provide decision auditing and human-in-the-loop participation Actionable Step You should adopt a layered agent strategy with a focus on: • Use MCP as your foundation • Combine framework-native tools for speed and meta-tools for complexity • Prioritize observability, scoped permissions, and secure input separation Consideration Agentic AI is powerful, but securing it is not just a technical requirement. It is an organizational responsibility now that requires clear ownership, principled design, and continuous validation.

1 week ago at the AWS New York Summit, Matt Wood launched a suite of new Generative AI features in Amazon Web Services (AWS) Bedrock. Rather than do an announcement post last week, I figured I'd use a few of them and give some more feedback in a week. Here's my take on the most impactful features: 1️⃣ Prompt Management - worth adopting immediately - replaces the janky prompt management approach you've probably hacked together with a well thought out managed service. Provides a standard, API-based (with a good UI) approach to manage prompts for your generative AI applications. I can see many tools standardizing / integrating with this over time. 2️⃣ Memory for Bedrock Agents - start POCs now - preview feature that's great for use cases where people interact with your AI app periodically over a long period of time. Provides an integrated approach for maintaining context over long interactions with an AI agent. 3️⃣ Bedrock Guardrails API - worth evaluating - lets you centralize and re-use evaluations of GenAI model inputs and outputs during development and in production. There are lots of good tools / libraries in this space today that handle the runtime, but this feature enables serverless execution. 4️⃣ Q Developer Customization - worth evaluating - lets Q understand your overall code base in addition to its native functions, so it can recommend calls into your proprietary libraries, and also utilize your organization's coding standards. Need to see it doing better at suggesting the right library function calls, but its definitely an improvement for large code bases 5️⃣ Code Interpreter for Bedrock Agents - limited POCs now - preview feature great for small tasks where deterministic results are important or to help with "last mile" data analysis. Lets your agent dynamically generate and run code (with no network access - so no API calls), but the limitations make it a 'niche' feature for now. A few that are in 'wait and see' mode: Prompt Flows - missing some key features, but could be very interesting combined with Agents once it's built out. Q Apps - need to understand why I wouldn't just use Bedrock Studio LLM Fine Tuning - cost performance improvements go to AWS & Anthropic. Only in one region and one model.

Amazon just partnered with OpenAI to build a Stateful Runtime Environment on Bedrock and I think this is the moment AI agents go from "cool demo" to "production infrastructure" in scale. The core idea is simple: AI agents need memory. Not just within a single conversation, but across sessions, across tools, across entire workflows. Right now most agent deployments are stateless, which means every time an agent starts a task it is starting from scratch. That is fine for a chatbot. It is not fine for an agent that is supposed to manage your billing reconciliation or handle customer onboarding flows. What Amazon and OpenAI are building is essentially the persistence layer that agents have been missing. The same way databases gave web applications the ability to remember things between requests, stateful runtimes will give AI agents the ability to remember things between tasks. For us at Flexprice, this is directly relevant. We are already building billing workflows that involve multiple steps across multiple systems, and the biggest technical challenge is maintaining context as an agent moves from one step to the next. A stateful runtime solves that at the infrastructure level instead of forcing every team to hack together their own persistence layer. If you are building anything with AI agents in production, this is the partnership to watch. The companies that figure out agent infrastructure early are going to have the same advantage that companies who adopted cloud infrastructure early had in the 2010s. What is the biggest pain point you have hit with AI agents in production?

𝐅𝐫𝐨𝐦 𝐎𝐧𝐛𝐨𝐚𝐫𝐝𝐢𝐧𝐠 𝐀𝐮𝐭𝐨𝐦𝐚𝐭𝐢𝐨𝐧 𝐭𝐨 𝐄𝐧𝐭𝐞𝐫𝐩𝐫𝐢𝐬𝐞 𝐒𝐜𝐚𝐥𝐞: 𝐋𝐞𝐬𝐬𝐨𝐧𝐬 𝐟𝐫𝐨𝐦 𝐁𝐮𝐢𝐥𝐝𝐢𝐧𝐠 𝐀𝐈 𝐀𝐠𝐞𝐧𝐭𝐬 𝐢𝐧 𝐧𝟖𝐧 𝐚𝐧𝐝 𝐀𝐦𝐚𝐳𝐨𝐧 𝐁𝐞𝐝𝐫𝐨𝐜𝐤. I built this automation workflow last week that reminded me of a workshop I recently led on creating Amazon Bedrock Agents and my first AI application with Amazon Q for Business. That comparison reminds me how powerful it is when context, automation, and data come together to build systems that actually think and act. I built this project for a client onboarding system. The goal was to simplify the entire process from the initial call to contract signing and post-onboarding communication and to see how far automation could go with minimal human intervention. Using n8n, I created an AI Agent that connects Google Sheets, Fathom AI, PandaDoc, Gmail, and Calendly into one seamless onboarding flow. Every time a deal closes, it:  • Extracts sales call transcripts using AI  • Generates contract fields automatically  • Sends PandaDoc for signing  • Updates CRM and triggers a welcome email Here’s what stood out when comparing this build to Amazon Bedrock Agents: 𝐒𝐢𝐦𝐩𝐥𝐢𝐜𝐢𝐭𝐲 𝐯𝐬 𝐒𝐜𝐚𝐥𝐚𝐛𝐢𝐥𝐢𝐭𝐲 n8n makes automation visual and fast. You can build and test agent logic in minutes. Amazon Bedrock requires setup across Lambda, IAM, and orchestration services but delivers scalable reliability and enterprise-grade fault tolerance. 𝐀𝐜𝐜𝐞𝐬𝐬𝐢𝐛𝐢𝐥𝐢𝐭𝐲 𝐯𝐬 𝐆𝐨𝐯𝐞𝐫𝐧𝐚𝐧𝐜𝐞 n8n is perfect for quick prototyping and cross-app integrations. Amazon Bedrock focuses on managed security, model access control, and compliance frameworks that support enterprise and regulated workloads. 𝐈𝐧𝐭𝐞𝐠𝐫𝐚𝐭𝐢𝐨𝐧 𝐒𝐩𝐞𝐞𝐝 𝐯𝐬 𝐄𝐜𝐨𝐬𝐲𝐬𝐭𝐞𝐦 𝐏𝐨𝐰𝐞𝐫 n8n connects instantly with Gmail, Docs, and CRMs using built-in nodes. Amazon Bedrock integrates deeply with S3, DynamoDB, SageMaker, and Amazon Q to create agent ecosystems that scale across organizations. 𝐒𝐞𝐜𝐮𝐫𝐢𝐭𝐲 𝐁𝐞𝐬𝐭 𝐏𝐫𝐚𝐜𝐭𝐢𝐜𝐞𝐬 𝐟𝐨𝐫 𝐧𝟖𝐧 If you build in n8n, treat your workflow like production code. ❤️🔥  • Store API keys and secrets in environment variables, not inside workflow nodes.  • Use n8n’s credential encryption for sensitive data.  • Limit user permissions and secure all webhooks with authentication.  • Rotate credentials regularly and review audit logs often. Both platforms aim to turn workflows into intelligent systems. n8n gets you there quickly. Amazon Bedrock keeps you there securely. If you want both speed and scale, build your prototype in n8n and operationalize it with Amazon Bedrock Agents.

Amazon Web Services (AWS) 𝗿𝗲𝗹𝗲𝗮𝘀𝗲𝗱 𝗮 𝗺𝗮𝘀𝘀𝗶𝘃𝗲 𝟴𝟬+ 𝗽𝗮𝗴𝗲 𝗴𝘂𝗶𝗱𝗲 𝗼𝗻 𝗛𝗢𝗪 𝘁𝗼 𝗯𝘂𝗶𝗹𝗱 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁𝘀 𝗶𝗻 𝗰𝗹𝗼𝘂𝗱-𝗻𝗮𝘁𝗶𝘃𝗲 𝘀𝘆𝘀𝘁𝗲𝗺𝘀. ⬇️ It reads like AWS’s vision for replacing traditional software stacks with autonomous, interoperable agentic systems. 𝗛𝗲𝗿𝗲’𝘀 𝘄𝗵𝗮𝘁 𝘁𝗵𝗲 𝗴𝘂𝗶𝗱𝗲 𝗰𝗼𝘃𝗲𝗿𝘀: ⬇️ → Frameworks like Strands, LangGraph, CrewAI, Bedrock Agents, and AutoGen — with implementation steps, use cases, and real-world deployments → Protocols like MCP and A2A — including how to choose the right one for enterprises, startups, and regulated sectors → Tooling strategy across protocol-based tools, framework-native tools, and meta-tools — covering memory systems, agent graphs, and workflow scaffolding → Security foundations including OAuth2.1, scoped permissions, sandboxing, audit trails, monitoring, and observability via CloudWatch and LangFuse → Implementation guidance — from evaluating frameworks to integrating tools, deploying across stacks, and scaling agents securely in production It's heavily centered around AWS-native services like Strands and Bedrock (who would’ve guessed) — but still an excellent read for technology leaders, architects, and developers who want to go beyond slideware and get hands-on with the actual frameworks, protocols, and implementation details. 𝗣.𝗦. 𝗜 𝗿𝗲𝗰𝗲𝗻𝘁𝗹𝘆 𝗹𝗮𝘂𝗻𝗰𝗵𝗲𝗱 𝗮 𝗻𝗲𝘄𝘀𝗹𝗲𝘁𝘁𝗲𝗿 𝘄𝗵𝗲𝗿𝗲 𝗜 𝘄𝗿𝗶𝘁𝗲 𝗮𝗯𝗼𝘂𝘁 𝗲𝘅𝗮𝗰𝘁𝗹𝘆 𝘁𝗵𝗲𝘀𝗲 𝘀𝗵𝗶𝗳𝘁𝘀 𝗲𝘃𝗲𝗿𝘆 𝘄𝗲𝗲𝗸 — 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁𝘀, 𝗲𝗺𝗲𝗿𝗴𝗶𝗻𝗴 𝘄𝗼𝗿𝗸𝗳𝗹𝗼𝘄𝘀, 𝗮𝗻𝗱 𝗵𝗼𝘄 𝘁𝗼 𝘀𝘁𝗮𝘆 𝗮𝗵𝗲𝗮𝗱 𝘄𝗵𝗶𝗹𝗲 𝗼𝘁𝗵𝗲𝗿𝘀 𝘄𝗮𝘁𝗰𝗵 𝗳𝗿𝗼𝗺 𝘁𝗵𝗲 𝘀𝗶𝗱𝗲𝗹𝗶𝗻𝗲𝘀. 𝗜𝘁’𝘀 𝗳𝗿𝗲𝗲, 𝗮𝗻𝗱 𝘆𝗼𝘂 𝗰𝗮𝗻 𝘀𝘂𝗯𝘀𝗰𝗿𝗶𝗯𝗲 𝗵𝗲𝗿𝗲: https://lnkd.in/dbf74Y9E

The pace of innovation led by generative AI is unprecedented. We’re seeing new use cases emerge across every industry that would not be possible without this technology. So, how can you help every developer build with GenAI in this rapidly changing environment? Here is the advice I shared during my keynote yesterday at #VivaTech:  🟠 Start your GenAI journey on Amazon Bedrock and give developers access to the broadest selection of first- and third-party LLMs and FMs from leading AI companies like Anthropic, Cohere, Meta, Mistral, and more. 🟠 Your organization’s data is the key differentiator between generic GenAI applications and those that know your business and customers deeply. Use enterprise data to customize foundation models and maximize their value. 🟠 Tackle repetitive coding tasks with Amazon Q Developer and adopt the use of autonomous agents to remove the heavy lifting from tasks like coding, writing tests, app upgrades, and security scanning. These assistants can also help employees use the right information to do their work better. 🟠 Build responsibly with safeguards for model outputs and receive model evaluation support with Guardrails for Amazon Bedrock. How teams invent today and tomorrow will have a profound impact on the world. That’s why we’re making generative AI accessible to customers of all sizes and technical abilities.