Aparna Sinha

People often ask me about the differences in architecture between Amazon Dynamo (the 2007 SOSP paper), DynamoDB (the AWS serverless NoSQL database), and Aurora DSQL (the AWS serverless SQL databases). I captured the key points on durability, high availability, programming model, and trade-offs in my blog: https://lnkd.in/gkMG_tVD TL;DR: Despite the shared name, Dynamo and DynamoDB have little in common architecturally, and make very different trade-off decisions.

631

8 Comments

Excited to give AWS customers access to Claude Sonnet 4.5. This latest model from Anthropic sets new standards for performance, especially for coding and building AI agents. Available today in Amazon Bedrock and coming soon to our Kiro IDE to help developers build even faster. https://lnkd.in/emAZSg92

6,534

128 Comments

As we move past the initial excitement of how to build and experiment with agents, we are now facing the classic day-two operations problems at enterprise scale: how do we meaningfully observe systems whose failure modes are more semantic than syntactic?     Traditional metrics like latency, error rates, and token counts feel like trying to debug a complex distributed system with only “ping”. They tell you if a service is alive, but not if it’s lucid. Businesses need to answer higher-order questions about factual accuracy and coherence.     Today at .conf, Splunk and AGNTCY | A Linux Foundation Project announced their collaboration to:   — Drive agentic semantic conventions directly into the OpenTelemetry (OTel) schema to establish a common language  —Integrate the AGNTCY Metrics Compute Engine (MCE) that provides the tooling, using an LLM-as-a-judge approach to evaluate qualitative signals like coherence alongside the operational metrics we already know. Consider a banking AI agent handling loan applications. Your operational dashboards might be all green, but the MCE could flag a persistent dip in "factual accuracy" for queries about variable APRs, revealing a silent, compliance-critical failure mode. Now you are no longer just monitoring the health of the container; you're monitoring the integrity of its reasoning.  This foundation becomes even more critical as we move from single agents to multi-agent systems, where specialized agents collaborate to execute complex jobs. Get your hands on the code. Join the workgroup meetings. Provide feedback.   Start with the open-source AGNTCY Observe toolkit on GitHub and instrument an agent here https://lnkd.in/gCUqVsff Photo from .conf today on the main stage highlighting where AGNTCY is delivering with agent observability

109

5 Comments

GPT-5 has arrived, and it's live across M365 Copilot, GitHub Copilot, Copilot, and Azure AI Foundry! It's the most powerful new LLM across several benchmarks and everything runs on Azure, with Foundry providing the coordination and security to support enterprise use. Azure AI Foundry's endpoint using model router allows developers to automatically direct their prompts to the most suitable version — whether it’s GPT-5, GPT-5-mini, GPT-5-nano, or GPT-5-chat. https://aka.ms/GPT-5-blog

70

4 Comments

→ MCP: Why It Matters More Than You Think → MCP Architecture  • MCP is a framework designed to enable different models to share and exchange contextual information efficiently. • The architecture focuses on interoperability - breaking silos of isolated AI models. • It consists of modular components that manage context data flow, synchronization, and version control. → Key Components That Make MCP Work • Context Manager: The brain that stores, updates, and retrieves shared contexts among models. • Communication Layer: The messenger, ensuring smooth data transfer between models over standard protocols. • Access Control Module: Maintains security and privacy by controlling who can read or modify context. • Sync Engine: Keeps context consistent and up-to-date across distributed systems. → Core MCP Architecture: The Backbone of AI Collaboration • Decentralized design enables scalability and robustness. • Extensible interfaces allow integration with diverse model types - from local LLMs to cloud APIs. • Real-time synchronization delivers up-to-the-moment context sharing for complex multi-model workflows. • Standards-based messaging fosters ecosystem-wide adoption and innovation. 🌟 Follow the AIKaDoctor (Free AI & Data Science Resources)  channel on WhatsApp: https://lnkd.in/dCTCEKKc 📌Follow Habib Shaikh For more such content.

415

37 Comments

Voice + AWS Bedrock & Claude + Strands Agents 100 people came to the AWS Builder Loft in San Francisco yesterday for a voice agents hands-on workshop. Natalie Rutgers talked about Deepgram's speech models and APIs. Ayan Ray gave a great overview of Amazon Web Services (AWS) AI models and services. Strands Agents is an open source agents SDK that works with any model/provider and has extensive AWS integrations. Strands will make as many tool calls as needed to execute a task, so in a conversational voice context you need to run the agent in a parallel process alongside the realtime voice pipeline. You also want to provide feedback about the progress the agent is making. Chad Bailey added <thinking> text output and a second voice to the voice-ui-kit debug console to show this code pattern. Thank you to Ayan, Eric Yuen, and AWS Partners for making this event possible. Code from the workshop is here: https://lnkd.in/gFGtJdmw Strands Agents homepage here: https://strandsagents.com/

45

6 Comments