AWS re:Invent 2025 - New Era of Platform Engineering – Agentic AI-Powered Self-Service (AIM359)

New Era of Platform Engineering – Agentic AI-Powered Self-Service

Overview of Platform Engineering

• Platform engineering evolved from DevOps, with the goal of creating internal development platforms or self-service platforms
• Allows organizations to standardize tools, processes, and practices across multiple development teams
• Helps avoid operational overhead and costs from inconsistent tool usage across teams
• Takes approximately 2 years for organizations to mature their platform engineering practices

Softsurf's Adaptive Modernization Platform

• Softsurf developed a framework rather than a proprietary platform solution
• Based on AWS best practices and customer feedback
• Includes core components like EKS, ECS, Lambda, observability, security, CI/CD, and an Internal Development Portal
• Designed to be easily customizable and allow substitution of individual components

Integrating Agentic AI into Platform Engineering

• In 2023, organizations began experimenting with generative AI like ChatGPT
• However, integrating AI into production environments proved challenging
• In 2025, the focus shifted to identifying business value and building a case for agentic AI

IDQ: AI-Driven Enhanced Engineering

• Softsurf's solution to integrate agentic AI into platform engineering
• Leverages first-party AWS agents, custom agents, reusable prompts, and large language models (LLMs)
• Allows developers to leverage AI capabilities without deep AI expertise
• Aims to address bottlenecks in the software development lifecycle (SDLC)

Real-World Use Cases

1. Automated Deployment Pipeline:

   • AI-powered orchestration of the CI/CD pipeline
   • Automatically detects programming language, builds container, deploys to EKS
   • Provides status updates and root cause analysis for deployment issues

2. Kubernetes Deployment Validation:

   • AI-driven validation of application and cluster components
   • Ensures deployments are validated before production release
   • Saves 3 months of development time compared to a custom tool

Demonstration: Cloud Migration with AI-Powered Self-Service

• Showcases a self-service migration tool that leverages AWS Transform and agentic AI
• Allows on-premises application owners to modernize their architecture without deep cloud expertise
• Automatically generates Docker files, Kubernetes manifests, and migration steps
• Enables replatforming and rearchitecture instead of a simple lift-and-shift approach
• Results in 2-3x improvements in migration and modernization timelines

Key Takeaways

• Platform engineering is a mature practice that organizations are rapidly adopting
• Integrating agentic AI into platform engineering can address bottlenecks in the SDLC
• AI-powered self-service capabilities can significantly accelerate cloud migration and modernization
• Softsurf's framework approach allows for customization and easy substitution of components
• Real-world use cases demonstrate tangible benefits in terms of time and cost savings