/Events /AgentCon - Miami /Schedule
AgentCon - Miami
Friday, June 12
Registration and Networking
Plenary ·Keynote: So You're An Agent Boss Now
A few years ago, generative AI felt like a party trick. Now it's in every workflow, every IDE, every meeting. But that was just the warmup. Agents are different. They don't wait for your prompt. They perceive, plan, and act toward a goal, sometimes through dozens of steps, across multiple systems, without you touching a thing. That's not an incremental improvement. It's a fundamentally different relationship between humans and software. In this keynote, we'll look at the full picture: where the agentic stack actually lives today (Azure AI Foundry, Semantic Kernel, AutoGen, MCP, low code tools and beyond), what's already reshaping real industries, and what your role looks like inside all of it. Agents need direction. They need judgment. They need someone who understands what good looks like and can tell when it's going sideways. That's you. The AgentBoss era isn't coming. It's here. The question is whether you're ready to lead it.
April Dunnam is a Principal Cloud Advocate at Microsoft, Microsoft Certified Trainer, and MVP alumni. Known for her ability to demystify the Power Platform and AI for both developers and makers, she blends deep technical knowledge with a knack for making complex topics approachable and fun. With a passion for community and sharing knowledge, April runs a popular YouTube channel dedicated to Power Platform, AI, and low-code development. She actively contributes to the Microsoft 365 and Power Platform community through sample contributions, community calls, and mentorship. April’s expertise has been recognized with the Microsoft Technology Excellence Award for her innovative Power Platform solutions. She is a frequent speaker at international conferences, helping businesses and developers harness the full potential of AI and the Power Platform. When she’s not advocating for the Power Platform, you can find her performing with her rock band or fueling her caffeine and LEGO building habits.
Break
Main Track ·AI Guardrails for Fintech: Securing Prompts, Documents, and High-Risk Payment Workflows
As fintech and payments rapidly adopt AI for underwriting, fraud detection, customer service, and compliance automation, they are also unintentionally creating new attack surfaces. Unlike traditional systems, AI models interact directly with prompts, unstructured data, and free-form user inputs—making them vulnerable to prompt injection, misleading instructions, hallucinated financial outputs, and unintended data exposure. The risk becomes even more pronounced in payments environments where AI touches invoices, receipts, IDs, bank statements, KYC documents, chargeback narratives, and merchant communications. This session demystifies the emerging risk landscape around LLM-powered financial workflows and provides a clear, practical approach to building AI guardrails that can operate in regulated, high-trust environments. We will break down how malicious inputs can manipulate model behavior, how unstructured documents can quietly leak sensitive data, and how weak retrieval pipelines can surface outdated or non-compliant policies. Attendees will learn a step-by-step framework for securing AI systems across the entire interaction lifecycle: • Prompt & Input Protection: sanitizing user messages, enforcing schema-based constraints, and preventing injection or misleading directives. • Unstructured Data Validation: verifying invoices, IDs, receipts, and contracts before the model reasons over them. • Semantic Guardrails: ensuring the model cannot generate unauthorized financial instructions, calculations, or compliance interpretations. • Policy-Aligned Outputs: mapping every model response to business rules, audit requirements, and regulatory boundaries. • Runtime & Governance Controls: red-team testing, retrieval isolation, PII minimization, logging, and human-in-the-loop escalation. Using examples from fraud operations, onboarding, chargebacks, dispute management, and merchant risk reviews, we’ll show how to blend technical safeguards with operational work flows so financial AI systems stay accurate, safe, and fully compliant. Attendees will leave with an actionable blueprint for deploying AI confidently across high-risk payments and fintech environments.
Divyarani Raghupatruni is Senior Director of Product, Data and Orchestration at Alacriti, where she leads product strategy for Stablecoins, faster payments, fraud prevention for real time rails. With 15+ years in fintech and payments, she spent six years at Block (Square) as Principal Product Manager, redesigning cart and checkout products, reporting and data platforms while establishing AI-first product practices. Her career includes building cross-border payments products at Transfast, which was acquired by Mastercard. Divya writes about the convergence of data, AI, and payments infrastructure on her Substack, focusing on stablecoins, tokenization, and agentic commerce.
Can We Make $1 Million in the Next Hour? (Live-Coding a Full-Stack App)
Can we make a $1 million? Probably not, but we can deploy a whole app! In this live-coding session, we will attempt the "impossible": going from a blank editor to a deployed, database-backed application with a functional UI before the session timer runs out.
This isn't a get-rich-quick scheme; This is a crash course on how to vibe code your way to success: we will use Claude and Gemini, NotebookLM to generate ideas, OpenArt for assets, and our favorite agents to generate our documentation on the fly.
While all can benefit from this session, it's designed for those who are skeptical or intimidated by agentic workflows, this session will cover:
The Product Manager Mindset: How to prompt for architecture, not just syntax. We will discuss the "Accept All" trap and how to spot hallucinations before they break production.
The Live Stack: A look at the specific toolchain for 2026 rapid deployment that makes this speed possible.
The Reality Check: We won't just show the happy path. We will do live debugging and discuss the security implications of shipping AI-generated code.
We'll crowdsource the app idea from a curated list at the start of the hour. Will we walk away millionaires? We can hope. Will you walk away knowing exactly how to build your next big idea? Definitely.
Chris Perrin is a tech geek, chaos wrangler, and all-around software leader with a knack for building high-performance teams and systems. With over 20 years of experience in software engineering, he’s helped teams—from the noblest of startups to giantest of enterprises—navigate the wild world of software development, Agile lifecycles, and getting more done. Chris loves demystifying complex tech, creating buzzwords and then turning buzzwords into real-world solutions. Whether he's talking about microservices, team performance, or trying to get AI to do his job, he brings energy, humor, and plenty of practical takeaways to every talk. When he’s not writing code or breaking things in the name of innovation, you’ll find him mentoring up-and-coming devs, writing decent-to-okay fiction, and geeking out over the latest in geek. Catch him at the conference, on LinkedIn, or wherever there’s strong coffee and great conversations!
Reinforcement Learning Brains
Generative AI is often "confidently wrong" because it lacks a feedback loop. Using the "Reinforcement Learning (RL) Brain" methodology, I will show how to build local "Reward Models" that train agents on your specific PR review standards. This talk moves from basic inference to Behavioral Fine-tuning, ensuring AI-generated code doesn't just "work," but aligns with your team's long-term architectural health.
A Senior Lead with 11 years in the software trenches and a Master’s in Data Analytics, I build intelligent systems that move beyond the hype and into production. My expertise lies at the intersection of Deep Learning and Neural Networks, where I currently pioneer "Reinforcement Learning Brain" methodologies and MCP architectures to scale engineering expertise. When I’m not fine-tuning models, I’m either perfecting my backhand on the table tennis court or navigating the backroads of America—having driven my Tesla through nearly every state, I’m as comfortable optimizing cross-country charging routes as I am architecting complex datasets.
Lunch and Learn: Scaling Responsible AI-From Enthusiasm to Execution
In her book, Scaling Responsible AI: From Enthusiasm to Execution, celebrated speaker, AI strategist, and tech visionary Noelle Russell delivers an exciting and fascinating new discussion of how to implement artificial intelligence responsibly, ethically, and profitably at your organization. Responsible AI promises immense opportunity, but unguided enthusiasm can unleash serious risks. Learn how to implement AI ethically and profitably at your company with Scaling Responsible AI. In this keynote, Noelle Russell reveals an executable framework to:
Harness AI’s full potential while safeguarding your firm’s reputation Mitigate bias, accuracy, privacy, and cybersecurity risks from the start Make informed choices by seeing through the hype and identifying true AI value Develop an ethical AI culture across teams and leadership Scaling Responsible AI equips executives, managers, and board members with the knowledge and responsibility to make smart AI decisions. Avoid compliance disasters, brand damage, or wasted resources on AI that fails to deliver. Implement artificial intelligence that drives profits, innovation, and competitive edge―the responsible way.
Noelle Russell is one of the most recognized voices on the future of artificial intelligence and responsible innovation. A multi-award-winning TEDx presenter and best-selling author of Scaling Responsible AI: From Enthusiasm to Execution, she has become a go-to resource for organizations seeking to harness AI’s power to drive growth while managing risk. She was named the #1 Agentic AI Leader in 2025, and she has taught more than 3.4 million learners through LinkedIn, equipping leaders and teams with practical tools to thrive in an AI-powered economy. With her blend of technical expertise and business acumen, she helps organizations unlock AI’s full potential to achieve scalable, measurable outcomes.
Lunch and Learn with Noelle Russell
Join us for a LUNCH and Learn session with Noelle Russell. In her book, Scaling Responsible AI: From Enthusiasm to Execution, celebrated speaker, AI strategist, and tech visionary Noelle Russell delivers an exciting and fascinating new discussion of how to implement artificial intelligence responsibly, ethically, and profitably at your organization. Responsible AI promises immense opportunity, but unguided enthusiasm can unleash serious risks. Learn how to implement AI ethically and profitably at your company with Scaling Responsible AI. In this keynote, Noelle Russell reveals an executable framework to:
Harness AI’s full potential while safeguarding your firm’s reputation Mitigate bias, accuracy, privacy, and cybersecurity risks from the start Make informed choices by seeing through the hype and identifying true AI value Develop an ethical AI culture across teams and leadership.
Securing MCP workflows for Agentic AI
Agentic AI and MCP have proliferated widely across the tech ecosystem in the last year. WIth this, comes a whole set of security concerns and issues. In this session we look at securing Agentic AI, MCP and the entire AI pipeline.
Muntaser is a researcher in AI and ML. Formerly a technical AI lead at Nvidia, he currently works at Insight Global as a Lead GenAI Engineer
Skill Issue: How to Write Skills That Actually Work
You've written a dozen skills. Some work, some don't, and you have no way to tell which. The agent says "you're absolutely right" while invoking the wrong one, and you keep re-explaining the same things to it. Without a way to measure what a skill adds, there's no way to find out. Meanwhile, the team next door is writing the same ones you already wrote, because nobody can find yours.
If you get why skills matter but can't get yours to do what you want, this is for you. 301, no "what is a skill," straight to the craft.
Skills are context artifacts: prose for flexible guidance, scripts for deterministic work, rules for hard constraints. Let's make them better:
A context artifact library that grows and patches itself: new skills emerge from agent friction, existing ones get fixed when they drift.
Design evals that grade only what the context adds — scenarios that probe the contribution, rubrics that ignore everything else, guards against state bleed and prompt leakage.
Pair every change with a second-model reviewer that catches regressions before merge. Version the library so rollback costs a checkout, not a postmortem.
Treat skills like code: scan, sign, gate at install. But also, treat them like prompts: add scanners for the attack surface conventional tooling can't see, like injection in the skill body, indirect poisoning through whatever the skill ingests, and tool-abuse paths that didn't exist before the agent had browsing tools.
Build a context artifact supply chain: registry, discovery, telemetry, staged rollout, so the team next door finds your skill instead of writing it for the third time. The same registry that solves discovery solves compliance: one push updates every agent in the org. Measure what proves reuse is real: installation and activation rates, because a skill nobody finds is a skill nobody uses.
Every practice above is itself a skill. "This is how we write context artifacts around here" belongs in the library: versioned, installed on every agent, graded by its own eval. The meta-skill has to earn its tokens too.
Baruch Sadogursky (@jbaruch) did Java before it had generics, DevOps before there was Docker, and DevRel before it had a name. He built DevRel at JFrog from a ten-person company through IPO, co-authored "Liquid Software" and "DevOps Tools for Java Developers," and is a Java Champion, Microsoft MVP, and CNCF Ambassador alumni. Today, he's obsessed with how AI agents actually write code. At Tessl, an AI agent enablement platform, Baruch focuses on context engineering, management, and sharing. On top of sharing context with AI agents, Baruch also shares knowledge with developers through blog posts, meetups, and conferences like DevNexus, QCon, Kubecon, and Devoxx, mostly about why vibecoding doesn't scale.
Guest Speaker
Plenary ·Break
Main Track ·Guest Speaker
Workshops ·Observability and security of hosted agents in Microsoft Foundry
AI enthusiast with extensive software engineering background. Veronika has experience in full-stack development using C#, .NET, Java and Typescript, Azure and AWS clouds and loves using AI to support all her ideas. International public speaker, hackathon volunteer, co-organizer of Boston Azure AI user group, and mentor. Veronika holds a master's degree in Information Technology. She likes dancing, traveling and practicing aerial fitness in her free time.
Who Let the Agent In? Securing MCP Servers in Production
What if your MCP server could confidently decide who gets access to what, without turning your codebase into a security nightmare? In this fast-paced session, we follow the journey of a simple MCP server as it evolves from an open endpoint into a fully secured, production-ready system. Along the way, you’ll see how authentication actually works in MCP, how to move beyond basic role checks into fine-grained, contextual authorization with OpenFGA, and how these pieces fit together in real-world scenarios. The highlight is a live demo where we lock down an MCP server step by step, making the invisible layers of security visible and practical. By the end, you won’t just understand MCP security, you’ll know exactly how to implement it or even offload it entirely so you can focus on building powerful agent-driven experiences.
Prachi Jamdade is a Developer Advocate at Gravitee, writing about APIs and AI. Prachi is making code talk through blogs and tech talks. She has worked with multiple startups and shipped global products. She is working on Developer Experience and Developer Education.