Proposal template

How PhonePe’s in-house Code Review platform (Arnold) helped control the Code Flood

One-line summary - in 1–2 sentences, describe the engineering problem, experience, or question your session explores.
We discuss how we designed and scaled PhonePe’s custom agentic code review platform. Moving away from fragile, chat-based wrappers we engineered an async, distributed pipeline that orchestrates micro-agents for effective code review, without removing the Human-in-the-loop. We also talk about some key design decisions while building the tool.

What problem are you addressing? What real engineering challenge, question, or experience is this session about?
As we navigate this era of frontier LLMs assisting all engineering functions in writing code, organizations expect this to unlock true speed. Enable engineers with AI tools and harnesses - we will naturally observe a massive uptick in code generation volume.

But now that everyone is a 10x engineer, why aren’t we shipping 10x products at 10x speed?

The ground reality however is that - while organizations have sped up the code generation part of the lifecycle, the bottleneck has simply shifted down the SDLC pipeline to human code review. A classic realization of Amdahl’s Law.

“Amdahl’s Law is a principle in computer science that calculates the theoretical maximum speedup of a system when only a portion of that system is improved. It dictates that the overall performance gain is ultimately limited by the sequential (un-improvable) fraction of the task, showing the law of diminishing returns”

This is not to suggest that we should haphazardly integrate AI and assume that total automation of the code review process will speed things up; maintaining a human-in-the-loop remains essential. We must now empower engineers with the right tooling that is effective at improving this speed.

Who is the intended audience?
Platform Engineering/Infrastructure/Developers

Level: intermediate/advanced

List one or two practical takeaways.

• What factors constitute effective code reviews, and are there ways to perfect this art.
• Can any org take on building such a platform
• Consider self hosting Arnold in your orgs.

What will you share?

• Why we went down this path
• How this platform was built
• Live demo
• Open-source tooling
• Before-and-after experience

What is your experience with this problem?

• Production system
• Internal engineering project
• Open-source project

What approaches failed, disappointed, or created unexpected problems?

• Justifying an investment of this stature was a hard problem.
• Assuming that a simple AI loop was enough to get the best result was incorrect.
• Using Gitlab runners to do the the actual review execution, was a challenge.

What will you do differently today?

• We discuss a few architectural decisions while building the platform, that could have shortened the delivery time

What trade-offs did you consider?
As a heavily audited payments company, PhonePe faced a classic build-vs-buy dilemma. Existing market solutions lacked affordable, air-gapped setups and the flexibility our teams needed. We chose to build a custom platform that empowers diverse engineering teams while meeting strict company-wide standards. The talk goes deeper on this aspect

How can this help other practitioners?

• Design decisions when building in-house platforms
• Applying similar principles of scaling a consumer product, how we improved adoption.

Current state:

• Production experience
• In progress
• Open source

#codereview #developerplatforms #design #platformengineering #gitlab #ai #agentic