Static documentation is the silent killer of microservice agility; if it isn’t derived directly from the code, it’s already a lie. In the race to scale, manual service registries inevitably fall out of sync with production reality, leading to architectural blindness and mounting operational risk. This session introduces a paradigm shift in technical asset management: the AI-Powered Service Catalog, an intelligent, self-documenting ecosystem driven by code, not manuals.

The talk will demonstrate how targeted file collection and a clear & concise instruction set enables the system to achieve accuracy in detecting project frameworks, dependency maps, API endpoints, and infrastructure resources like Kubernetes and Azure Key Vault.

What You Will Learn

• AI driven Metadata Extraction: Using Al, to perform static code analysis and interpret/monitor the source code to generate structured YAML metadata that strictly follows internal CMDB schemas.
• Smart Merge Logic: Implementing a deep-merge strategy that distinguishes between static business fields (Cost Center, Ownership) and dynamic technical facts (Frameworks, API Deps).
• GitOps-Driven Updates: How automated triggers—such as merging a PR—activate GitHub Actions to refresh the catalog whenever new packages are added or frameworks are upgraded.
• Reducing "Noise": A targeted file collection strategy that minimizes context window usage and optimizes AI analysis. For ex: For .NET projects, using NuGet packages to achieve specific outcomes like circuit breaker status or database dependencies.

Who Should Attend

• Platform Engineers tasked with building and maintaining internal developer portals and service registries.
• Software Architects looking to solve the problem of architectural drift and undocumented service dependencies.
• DevOps Leaders interested in integrating AI-powered static analysis into their existing CI/CD and GitOps pipelines.
• Engineering Managers seeking to reduce developer toil and ensure accurate data for compliance and reporting.