How I Refactored a Legacy Codebase with Ralph That Was Estimated to Cost 1 Dev Month

Ralph is a technique coined by Geoffrey Huntley(https://ghuntley.com/ralph/) to run coding agents in a Bash Loop. This technique allows AI agents to run for as long as is required for them to complete a task, checking their work as they go. This is a breakthrough for AI agents to complete long running tasks.
I will demonstrate how I was able to refactor a production code base that would have taken a developer a month otherwise with just coding agents

Takeaways
=========
- You will learn about the typical loop for running a Ralph Wiggum (https://ghuntley.com/ralph/))
- You will learn the tips and tricks to make Ralph work on a production code base

PRO WORKSHOP: How I Refactored a Legacy Codebase with Ralph That Was Estimated to Cost 1 Dev Month

Ralph is a technique coined by Geoffrey Huntley(https://ghuntley.com/ralph/) to run coding agents in a Bash Loop. This technique allows AI agents to run for as long as is required for them to complete a task, checking their work as they go. This is a breakthrough for AI agents to complete long running tasks.
I will demonstrate how I was able to refactor a production code base that would have taken a developer a month otherwise with just coding agents

Takeaways
=========
- You will learn about the typical loop for running a Ralph Wiggum (https://ghuntley.com/ralph/))
- You will learn the tips and tricks to make Ralph work on a production code base

Dev Roundtable: Tuhin Kanti Sharma, Principal Software Engineer, Development Experience

From Legacy to AI-Native: Scaling Development with the Ralph Loop

In this roundtable, I will share key lessons learned from leading a team through the transformation of a legacy enterprise product into an AI-native solution. We will explore the specific personas that utilize agents most effectively and discuss the growing hurdles faced by developers who have yet to integrate AI into their workflows.

I will provide a deep dive into my personal workflow, specifically how I use the Ralph Loop to scale development and why AI has become essential to my success. Finally, I will share practical strategies for managing token budgets to maximize the value of your allocated quota.

Serverless Deployment Pipeline using Packer & Terraform

We will like to demonstrate our serverless cloud native architecture to build a continuous deployment pipeline to build 100s of environments everyday for integration testing. The pipeline is composed of services that build upon an on demand Packer service to package the applications and an on demand Terraform service to provision the resources. The infrastructure scales and refreshes on demand and enables us to continuously deploy our applications. These applications are a set of interdependent micro-services that our platform is able to service.

This session is targeted for engineers looking to build continuous deployment systems on public cloud using terraform and packer