Federated Agents and Distributed Agentic AI
The most valuable data is the data that never leaves the system. Information Silos and legacy system complexity offer the highest friction to the Adoption of Agentic AI in most businesses. While many teams attempt to retrofit an MCP layer into their existing systems and workflows, it solves very little.
In this talk, we will understand what a Federated Agentic model looks like. We will delve into the strengths of having agents that own specific components and interests, as opposed to equipping a generalist agent with tools to operate the whole system. We will also explore why Agentic AI is fundamentally a Distributed Systems problem, and how tools and concepts from the Distributed Systems paradigm apply to Federated Agentic AI.
You will leave with an understanding of how to build Localized, Specialized, and Efficient AI Agents that run where the data lives, serve the interests of the teams that own them, and still operate securely in a system that potentially spans the globe (or even space ;) )
NATS: Distributed Systems made Simple
In this talk, I aim to draw focus to NATS, particularly NATS Jetstream. This project, in my opinion, have been overshadowed by the mainstream heavyweights like Kafka, RabbitMQ, etc.
The following are a few touch points we will go over:
1. A quick overview of NATS and NATS Jetstream
2. Comparisons with Kafka, RabbitMQ, etc.
3. How it's used currently vs. How it can be used (Shout out to NATSbyExample)
4. Highlighting why NATS is a game changer for distributed systems and Microservices:
- Built in service discovery, KV, Object Storage
- Auth, Supercluster, leaf nodes, etc
- NATS execution (NEX)
- Polyglot services setup
5. A quick demo project leveraging most of the features above.
IoT Edge Insights Made Simple: Unleashing OpenTelemetry's Power
IoT and Distributed Systems come with numerous challenges of their own. Here, we will attempt to use OpenTelemetry and concepts from the Kubernetes and microservices Observability paradigm to the field of IoT. In doing so, we will achieve better observability and understanding of IoT systems and their components. Such understanding of the functioning of an IoT system will allow for deeper insights, easier triage and resolution of issues and better capacity planning. The marriage of OpenTelemetry and Kubernetes also extends to actionable insights, focusing on curating metrics, logs, and traces, best practices for meaningful alerting, queries, and operational dashboards, and addressing interoperability and tooling challenges in the IoT landscape. We hope to simplify the practical implementation of OpenTelemetry's instrumentation libraries, enabling developers to seamlessly integrate observability in their IoT applications.
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top