Áron Erdélyi
Senior Consultant @ TNG | Full-Stack Developer | Generative AI Specialist
Austin, Texas, United States
Actions
Aron Erdelyi is a Senior Consultant at TNG Technology Consulting, where he focuses on building AI-enabled systems and integrating modern AI capabilities into scalable, cloud-native architectures. His work spans generative AI applications, intelligent microservices, and practical AI tooling for developers. At TNG, he helps teams adopt AI responsibly while keeping engineering quality, performance, and maintainability front and center. Outside of work, Aron is an accomplished rugby player. He has earned 14 caps for his national team and captains the reigning national champions.
Area of Expertise
Topics
What the hell is up with MCP?
MCP has rapidly become the de facto standard for tool calling in AI systems - not because it was perfectly designed, but because it was the first open protocol to gain real traction. In this talk, I detail how MCP evolved from an early standard to a widely adopted foundation across major LLM providers and thousands of public servers.
With explosive growth comes serious challenges: fragmented implementations, unresolved security risks such as tool poisoning and prompt injection, and context bloat as servers accumulate large tool definitions. The ecosystem continues to shift quickly - new features like asynchronous operations and programmatic tool calling have landed, and as of December 2025, MCP has moved under the Linux Foundation’s Agentic AI Foundation, bringing neutral governance from major industry contributors.
We will cut through the complexity to show what the protocol truly specifies, how it’s used in practice, and which features matter to build robust clients and servers as MCP enters its next phase.
Break me if you can: Evaluating safety robustness in AI models
AI safety is often framed as an ethical question, but for businesses shipping real products, it is also a matter of security, trust, and legal integrity. Unsafe model behavior can expose secrets, amplify bias, enable abuse, and create serious reputational risk. This talk gives developers and product teams a practical introduction to AI alignment and AI safety through the lens of how systems actually fail in production.
From there, we examine how the AI red teaming field is moving from manual probing to automated adversarial evaluation at scale, including static test sets, agentic jailbreaking, and optimizer-driven attack discovery. They close with a practical threat model, concrete mitigations, and an overview of the current AI security landscape, giving developers a grounded framework for building safer, more secure AI systems.
MCP and the Emerging Agent Protocol Stack
_A field guide to MCP, A2A, generative UI, and related standards._
MCP has rapidly become the de facto standard for tool calling in AI systems, not because it was perfectly designed, but because it was the first open protocol to achieve real traction. What began as a pragmatic interface for connecting models to tools has now become a foundation for a much broader ecosystem.
We trace how MCP evolved from an early standard into widely adopted infrastructure across major LLM providers and thousands of public servers, and we examine the challenges that came with that growth: fragmented implementations, unresolved security risks such as prompt injection and tool poisoning, and increasing context bloat as servers expose ever-larger tool surfaces.
But tool calling is no longer the whole story. As AI systems become more capable, they also need richer interfaces, longer-running workflows, asynchronous communication, and ways to share artifacts across systems. That pressure has driven the emergence of adjacent ideas and protocols: MCP Apps and A2UI for generative user interfaces, and A2A for long-running, multi-agent interactions.
We cut through the noise to explain what these agentic protocols actually specify, what they don't, and how it fits into this large AI landscape. The result is a practical map of where these protocols stands today and how these efforts are shaping the next generation of agentic systems.
BigTechDay 26
I will be hosting 2 talks:
- Break me if you can: Evaluating safety robustness in AI models
- MCP and the Emerging Agent Protocol Stack
also I will exhibit an AI Red Teaming platform
DevDays Europe 2026
I will be hosting the session: 'What the hell is up with MCP?'
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