Demo Theater

Transforming AI inference from a centralized bottleneck into an intelligent, globally distributed compute cache that instantly adapts to user demand.

Presenters:

• Duan Lightfoot — Senior Developer Advocate, Akamai

View in the Schedule

How do you stop an AI agent from leaking API Keys, OAuth tokens, or secret credentials? “Hey, here is my API key, can you do this for me” an AI agent can easily ask another agent/API/MCP server. The answer is simple. Don’t give them any credentials. So how do they do any meaningful work then? Inject the credentials on egress. In this quick session, we’ll look at credential brokering for AI agents and how agent gateways can be used for this.

Presenters:

• Christian Posta — Global Field CTO, Solo.io

View in the Schedule

Bring your app’s interactive features directly into ChatGPT. See how plugin extensions power an interactive design workspace where users can browse designs, edit a canvas, and guide changes through conversation, selections, and annotations. We’ll use MagicPath, a design startup that saw a 2,000% increase in new users after launching its ChatGPT plugin, as an example. See its interactive workspace in action, then learn what founder Pietro Schirano took away from the launch—with practical lessons for building and growing your own ChatGPT plugin.

Presenters:

• Ted Ramert — Applied AI Engineer, OpenAI

View in the Schedule

See Device Connect in action as an AI agent discovers and interacts with real-world devices. We’ll demo the end-to-end flow live, showing how developers can connect agents to device capabilities and extend agentic experiences beyond the cloud and into the physical world.

• Kavya Sri Chennoju — Staff AI Engineer, Arm.

View in the Schedule

If you expose your APIs as MCP tools one endpoint at a time, you leave it to the agent to run a multi-step business process on its own. It decides the order of calls. It has to handle partial failures. It might skip a step you assumed was required.

This session shows a different way to build MCP tools. A composite tool wraps an entire multi-step process, including validation and rollback, into one action the agent calls once instead of five. We will build one live. A multi-step order fulfillment process becomes a single governed MCP tool, with every call tied to the real user making the request instead of a shared service account.

If you are building MCP servers by wrapping endpoints one at a time, this talk shows the problem with that approach and a way to fix it.

Presenters:

• Tyler Lamparter — Product Portfolio Strategist, Celigo

View in the Schedule

An AI gateway is only as performant as the data layer powering it. Join Jerod Johnson, CData’s Director of Technology Evangelism, for a live walk through of the Connect AI Gateway, built to govern what agents do at the data layer, and ground them in context that spans the organization.

Presenters:

• Jerod Johnson — Director, Technology Evangelism, CData Software

View in the Schedule

See how Workato lets you build smarter, more reliable, and more token-efficient MCP servers by leveraging structured tool calls that rely on deterministic logic rather than probabilistic intent.

Presenters:

• Alex Stewart — Technical Advocate, Workato

View in the Schedule

See Smartsheet’s AI-powered platform become the orchestration layer for your organization’s work. We’ll build a real-time operational dashboard live with Smartsheet AI: standardized reporting, exception-based alerts, and views that leadership trust.

Presenters:

• Brian Gann — Principal Solutions Engineer, Smartsheet

View in the Schedule

Your teams are already building with AI, but most of that work is stuck in one-off agents, personal prompts and disconnected tools that never get past the person who built them. Agents, MCPs and skills keep multiplying across departments. There’s no shared catalog, no view into token spend and no way to prove which workflows deliver ROI. In this session, you’ll see how Runlayer turns that sprawl into a single catalog of agents, skills and connectors that any employee can use in Slack, with usage, spend and performance insights built in. AI transformation leaders and builders will leave knowing more clearly how to scale AI adoption across the workforce and show that it’s paying off.

Presenters:

• Keith Ledgerwood — Forward Deployed Engineer, Runlayer

View in the Schedule

This session will show you how to use Agent substrate and AX to scale your agents.

Presenters:

• Ivan Nardini — AI Dev Engineer, Google

View in the Schedule

Porting a frontier model to custom silicon requires deep Neuron expertise, days of iteration, and hard-won knowledge that rarely gets codified. Neuron Agentic Development automates this journey from Hugging Face model to optimized model running on AWS Trainium, AWS’ custom AI accelerator used by leading providers. In this session, we’ll demo using agentic loops to port a model, verify equivalence, and run optimization loops. We’ll also share learnings from building loop engineering in specialized domains.

Presenters:

• Mike Chambers — Senior AI Specialist DA, AWS

View in the Schedule

Tokenomics covers the full path from energy and capital into AI, efficiently consuming it, and delivering measurable business outcomes. That path has three stages, and the Tokenomics Foundation is building an open framework to map this new territory. These are the lessons that came out of the working sessions and discussions behind that framework, from the people building and living these problems. Open, neutral, and community-built under the Linux Foundation.

Presenters:

• Stephen Arthur — Member of Technical Staff, Tokenomics Foundation

View in the Schedule

AI agents are getting good at taking action. The hard part isn’t making them capable — it’s knowing when to make them stop.

This talk walks the full spectrum of human-in-the-loop patterns: inline confirmations, out-of-band permission gates, and handing your agent a wallet with real money in it. Each fits a different level of consequence. Built around live demos that escalate in stakes with every step, you’ll leave with a mental model and a reference architecture you can use the same day.

Presenters:

• Michael Liendo — Staff Developer Advocate, Okta

View in the Schedule

Your agents keep working when you leave your desk. We’ll show what we’re doing so you can keep up with them on mobile, and share the challenges of building open source iOS and Android apps.

Presenters:

• Tom Brow — Software Engineer, Block
• Kenny Lopez — Product Designer, Block

View in the Schedule

Every organization has workflows buried in back-end systems – work orders, field reporting, asset tracking. They work fine. But AI agents can’t reach them. This demo shows what changes when they can.

The remarkable part: the underlying workflow already existed. It was built years ago. The only new step was exposing it to the AI, and that took under seven minutes – with no code written by anyone.

For enterprise, the message is simple. You don’t need to rebuild anything. Your business logic, your data rules, your integrations – they’re already done.
The missing piece was a way for AI to call them safely. That piece now exists, and getting there requires neither a developer nor any knowledge of the technology underneath.

The question isn’t whether your operations are AI-ready. It’s how quickly you can flip the switch.

Presenters:

• Matt Meeboer — Customer Solutions Specialist, Safe Software

View in the Schedule

When people talk about “open source” in today’s AI & ML landscape, they likely think of open models and kick off another debate between “open weights” and truly “open source” models. But open source goes beyond models… it’s behind the most well-known and critical pieces of the ecosystem, from training to chunking data & embedding for RAG to inference to scaling, evals, observability, and optimization. Open source is everywhere and driving much of the AI ecosystem. For this session, we invite technical practitioners and engineering leaders to join Red Hat and open source expert Wesley Chun to learn what some of these critical pieces are, including but not limited to: PyTorch, vLLM, llm-d, OpenShell, OGX (formerly Llama Stack), Docling, and others.

Presenters:

• Wesley Chun — Technical Program Manager – AI, Red Hat

View in the Schedule

Learn how to deploy and operate a resilient, air-gapped AI agent platform tailored for harsh industrial environments, including smart factories, offshore oil rigs, and cargo ships. This session demonstrates a production-grade, open-source stack built on Red Hat, Kubernetes, ArgoCD, Keycloak, APISIX, Temporal, and Hitachi Agentic Runtime designed for 24/7 operational continuity without internet connectivity. Discover how to run, orchestrate, and observe stateful LangGraph and Agent Development Kit (ADK) agents directly on the plant floor and remote edge locations with zero telemetry leaks and industrial-grade reliability.

Presenters:

• Georgy Okrokvertskhov — Vice President, Engineering, GlobalLogic

View in the Schedule

See how organizations can bring trusted, practitioner expertise directly into AI workflows like Claude, Copilot, and coding environments. We’ll demonstrate how grounding AI in authoritative sources improves technical decisions, code quality, and confidence without slowing teams down.

Presenters:

• Ro Recio — Sales Engineer, OReilly

View in the Schedule

One agent invocation fans out into dozens of model calls and tool calls, across shared inference, shared retrieval, and shared infrastructure. By the time it shows up on a bill it is one anonymous number billed to whoever provisioned the cluster. Tags cannot decompose it, and a gateway only sees the traffic that went through the gateway. This session shows a different approach: read the runtime instead of the invoice. We will put a live agent workload on screen and break a single shared pod down to the team, feature, and customer that actually caused the spend, with no tags, no SDKs, and no code changes.

Presenters:

• Amit Kinha — Field CTO, DolT

View in the Schedule

Many agentic AI projects risk cancellation due to spiraling inference costs and unmanaged execution. Connecting AI agents to tools via MCP enables powerful workflows, but ungoverned access leads to tool bloat, context rot, and runaway token consumption. In this session, we’ll discuss how Nutanix Agent Gateway – built by maintainers of the AgentRouter open source project – establishes unified runtime control across MCP tool servers and LLM endpoints for enterprises. Learn how to eliminate credential sprawl, enforce function-level tool filtering using MCP Connectors, and apply token rate limiting to prevent bill shock while balancing public and private models.

Presenters:

• Ashwini Vasanth — Group Product Manager for Enterprise AI, Nutanix
• Laura Jordana — Principal Technical Marketing Engineer, Nutanix

View in the Schedule

Metering used to mean counting a massive stream of a single event type, but agents turn it into a horizontal problem spanning models, tokenizers, entities, entitlements, organizational hierarchies, and constantly changing pricing models. I encountered this firsthand while building a decentralized metering system with pluggable measurement components, contract and organization hierarchies, entitlement-aware usage, and built-in data reconciliation.

This talk distills that architecture into a generic model for agentic metering, separating observation, measurement, attribution, and pricing so each can evolve independently.

Presenters:

• Shai Betito — VP of Engineering, Stigg.io

View in the Schedule

What actually happens when an AI agent tries to get something done on your website? In this 10-minute demo, we’ll spawn real AI agents live, running on different harnesses and models, and give them the same task on an agent-ready website and on one that isn’t. You’ll see in real time where they succeed, where they get stuck, and how much agent readiness changes the outcome. Along the way, we’ll show how ora scans websites for agent readiness and walk through the quality metrics we use to score every agent journey.

Presenters:

• Ido Finder — Founding AI Lead, Ora

View in the Schedule

Guardrails live inside the model, and a model can be talked out of them. This demo puts enforcement in the runtime instead, a single control plane that every agent tool call passes through. Your identity system and your own policies decide what each user’s agent can see and do, and the runtime applies those rules and logs every call it allows or blocks. Watch it hold a wire transfer for human approval, strip customer PII from a response before the model reads it, and shut down a user who borrowed a colleague’s connection to reach tools they were never granted.

Presenters:

• Shub Argha — Head of Forward Deployed Engineering, Arcade.dev

View in the Schedule

Most developers already use multiple AI agents. The challenge is getting them to work together without constantly managing context and handoffs. Our five developers now work alongside 45 agents, tackling real projects, bug fixes and code reviews. Different agents, different models, one shared context that doesn’t belong to any AI provider. Join us for a 10-minute live demo of Switch AI. We’ll show you how our team works, the use cases we’ve built, and how to get your own agents working together without being locked into one tool or model. Switch AI is free and source available, bringing any agent, from any framework, into the tools your team already uses. https://github.com/sandbox-quantum/switch

Presenters:

• Phil Conway — Developer Advocate, SandboxAQ

View in the Schedule

Most agent memory does not need a graph. If a saved value only has to point back to its source message, an ID could be sufficient. But when the same entities appear in different conversations and connect to other parts of the agent’s history. investigating that history is no longer as simple as following a single predefined lookup. It means traversing through the connections in memory from whichever part of the system raised the question.

In the live demo, I’ll connect an agent to an open source agent memory workspace through MCP and retrieve context from conversations outside its current session. I’ll then inspect the resulting context graph, following the connections between the retrieved entities and the earlier messages in which they appeared. Along the way, I’ll distinguish the links agent memory workspace creates during extraction from the provenance an application has to record deliberately.

Presenters:

• Nyah Macklin — Sr. Developer Advocate, Artificial Intelligence, Neo4j

View in the Schedule

Building efficient agents requires understanding how model capabilities, agent frameworks, and tools work together. This session introduces Step 5 Preview and explores practical considerations for model selection and agent design. We’ll discuss how to balance task quality and cost across coding and professional workflows, with an emphasis on efficiency at the task level.

Presenters:

• Frank Li — Head of Global Developer Relations, Stepfun

View in the Schedule

AI agents need tools to act, and enterprises have spent years on tailoring APIs to gain access to enterprise data. The challenge isn’t simply exposing them through MCP. It’s deciding which capabilities agents should see, how they should be composed, and how access is governed when agents start acting on enterprise data. In this session, we’ll follow a simple use case from API to agent. We’ll transform existing REST APIs into reusable MCP Tools, compose them into a purpose-built MCP Server, and make those tools available to an AI agent. Then we’ll add the controls needed for the real world: agent identity, MCP server access registration, delegated human access, and runtime authorization over the data behind each tool. You’ll see how API management, MCP, identity, and authorization come together to create a governed Agent Tool Layer, turning the APIs you already have into secure, composable capabilities built for agents.

Presenters:

• Cruse Sweeney — Solutions Engineer, Gravitee

View in the Schedule

The days of tokenmaxxing are over. Join Kevin from PointFive to see how TokenShift helps organizations attribute how developers are truly spending their tokens, apply controls to govern usage, and even optimize activity to improve AI efficiency.

Presenters:

• Kevin Rhine — Founding Solutions Engineer, PointFive

View in the Schedule

A live walkthrough of how to build, submit and improve plugins for Claude users. See how to package skills, connectors and commands into a plugin, get it listed in the Claude directory, and best practices for keeping it up to date.

Presenters:

• Tobin South — Member of Technical Staff, Anthropic

View in the Schedule

See how Kong’s unified API and AI platform takes agents from experiment to production: governing a real multi-agent system with A2A communication, MCP tool calls, and multi-provider LLM routing, all with end-to-end observability across the full data path.

Presenters:

• Deirdre Anderson — Solutions Engineer, Kong

View in the Schedule

What happens when your AI agent or MCP server goes down mid-task? What about when the API or database it depends on starts dropping connections? Your agent probably won’t give up, because LLMs are persistent. But chances are high that it will redo everything over and over. Re-running tool calls, re-spending tokens, even for work that already finished. You’ll still get your result, just slower, and at a higher cost.

In this live demo, we’ll show what changes with Durable Execution. You’ll join in the fun to become the chaos that breaks our demo. We’ll watch the full execution path (Agent, MCP server, APIs and storage) while Temporal keeps resuming from where things left off, not starting over.

You’ll walk away knowing how durable execution works in both agent code and MCP server code, how retries and replays prevent unnecessary rework and token spend when services fail and recover. Come see how “what if this other system goes down?” stops being a question you need to build custom handling for within your code.

Presenters:

• W. Ian Douglas — Staff Dev Advocate, Temporal
• Melissa Herrera — Senior Developer Advocate, Temporal

View in the Schedule

An agent can have the right credentials, stay inside its sandbox, and still do the wrong thing.

Give an agent a box and, like a gas, it will expand to fill every dark corner. The walls define where it CAN go, not where you WANT it to go.

In this live demo, we’ll give an off-the-shelf agent a task and access to ordinary MCP servers. Then we’ll explore the gap between what it can do and what we actually intended. See intent-based authorization at work, without replacing the agent or rewriting the tools.

Presenters:

• Geoff Goodman — Member of Technical Staff, Keycard

View in the Schedule

Inference, GPUs, networking, storage and scale all matter. Come see the infrastructure behind production AI workloads with Vultr.

Presenters:

• Mayank Debnath – Director of Developer Relations, Vultr

View in the Schedule

Taking agents to production means proving controls are in place. That’s where teams get stuck. Join us for a 10-minute live demo of WSO2 Agent Manager and see how trusted AI governance lets you monitor what your agents are doing, prove they are working, and shut one down if you need to.

Presenters:

• Tori Jeffcoat — Director, Product Marketing | AI, WSO2
• Muhammad Ali — Lead Sales Engineer, WSO2

View in the Schedule

A look at the future of agent authorization beyond coarse OAuth scopes. A live demo of agent intent governance and where nondeterminism belongs in authz.

Presenters:

• Aaron Tainter — Software Engineer, WorkOS

View in the Schedule