Barry Stahl
Architect & Developer focused on building reliable, self‑healing systems
📋 Biography
Barry Stahl is a Solution Architect and long‑time engineering mentor with more than four decades of experience building reliable, scalable systems on Microsoft technologies. He speaks at conferences and user groups on architecture, problem‑solving modalities, and the practical mechanics of building resilient software. Barry is known for making complex topics clear, actionable, and grounded in real‑world engineering practice. Based in Phoenix, he helps developers understand when deterministic or stochastic approaches to AI are appropriate, when AI is not the right tool, and how to design systems that stay reliable, self‑healing, and resilient under real‑world pressures.
✨ High-Impact Contributions 5
Genetic algorithms "learn" to make better decisions by making continuous improvements in strategy based the fitness of that solution for survival. These algorithms, modeled after Darwinian evolution, can solve complex optimization problems across many domains - from resource allocation to network design to automated testing. In this talk we'll define the DNA of our solutions, explore how to represent different types of problems in genetic terms, and examine the parameters that control how solutions evolve and improve. You'll leave with practical knowledge of how to apply these powerful techniques to your own challenging problems.
From genetic algorithms to ant colonies, nature has inspired some of the most powerful problem-solving techniques in computer science. This comprehensive workshop takes you on a journey through three interconnected approaches to algorithmic problem-solving: Dynamic Programming as a foundational technique for optimal decision-making, Genetic Algorithms for evolving solutions to complex optimization problems, and Bio-inspired Algorithms like Firefly and Ant Colony for tackling challenging search spaces.
Large Language Models (LLMs) have revolutionized the field of natural language processing, enabling machines to generate human-like text with remarkable coherence. This workshop offers a deep dive into three pillars of these architectures: Transformer models, Embeddings, and Retrieval Augmented Generation (RAG). We’ll start by unpacking transformers, including attention mechanisms and the token prediction process to understand their strengths and limits. Next, we'll explore embeddings and techniques like cosine similarity and clustering to build advanced applications beyond chatbots. Finally, we'll discuss implementations of RAG including both text-based and graph-based methods that address real-world challenges. You'll gain strategies for retrieval accuracy, bias handling, and response validation. By the end, you’ll master both theory and practice to responsibly harness LLMs for content generation and information retrieval.
This workshop teaches .NET Aspire through hands-on development of a multi-service application. You'll leverage a browser-based frontend communicating with .NET backend services and both SQL and NoSQL databases. The project leverages industry-standard OpenTelemetry for distributed tracing and metrics, configures service discovery for inter-service communication, and uses .NET Aspire's dashboard for monitoring and service health during development. You'll work with containerized dependencies, replace manual connection strings with Aspire's integration providers, and add .NET Aspire's features into an existing application. Additionally, attendees will craft deployments to Azure Container Apps and Kubernetes, leveraging IDE tools and GitHub Actions workflows for CI/CD automation. Attendees leave with a GitHub repo with extensive samples and the courseware to continue learning or apply the knowledge in their own projects.
Presented "LLM Embedding - Not Magic, Just Math" to attendees at MLCon NY 2025. The presentation was also recorded and is available on the Devmio platform.