How Nubank refactors millions of lines of code to improve engineering efficiency with Devin

8x
engineering time efficiency gain
20x
cost savings
Vimeo

Overview

One of Nubank’s most critical, company-wide projects for 2023-2024 was a migration of their core ETL — an 8 year old, multi-million lines of code monolith — to sub-modules. To handle such a large refactor, their only option was a multi-year effort that distributed repetitive refactoring work across over one thousand of their engineers. With Devin, however, this changed: engineers were able to delegate Devin to handle their migrations and achieve a 12x efficiency improvement in terms of engineering hours saved, and over 20x cost savings. Among others, Data, Collections, and Risk business units verified and completed their migrations in weeks instead of months or years.

The Problem

Nubank was born into the tradition of centralized ETL FinServ architectures. To date, the monolith architecture had worked well for Nubank — it enabled the developer autonomy and flexibility that carried them through their hypergrowth phases. After 8 years, however, Nubank’s sheer volume of customer growth, as well as geographic and product expansion beyond their original credit card business, led to an entangled, behemoth ETL with countless cross-dependencies and no clear path to continuing to scale.

For Nubankers, business critical data transformations started taking increasingly long to run, with chains of dependencies as deep as 70 and insufficient formal agreements on who was responsible for maintaining what. As the company continued to grow, it became clear that the ETL would be a primary bottleneck to scale.

Nubank concluded that there was an urgent need to split up their monolithic ETL repository, amassing over 6 million lines of code, into smaller, more flexible sub-modules.

Nubank’s code migration was filled with the monotonous, repetitive work that engineers dread. Moving each data class implementation from one architecture to another while tracing imports correctly, performing multiple delicate refactoring steps, and accounting for any number of edge cases was highly tedious, even to do just once or twice. At Nubank’s scale, however, the total migration scope involved more than 1,000 engineers moving ~100,000 data class implementations over an expected timeline of 18 months.

In a world where engineering resources are scarce, such large-scale migrations and modernizations become massively expensive, time-consuming projects that distract from any engineering team’s core mission: building better products for customers. Unfortunately, this is the reality for many of the world’s largest organizations.

The Decision: an army of Devins to tackle subtasks in parallel

At project outset in 2023, Nubank had no choice but to rely on their engineers to perform code changes manually. Migrating one data class was a highly discretionary task, with multiple variations, edge cases, and ad hoc decision-making — far too complex to be scriptable, but high-volume enough to be a significant manual effort.

Within weeks of Devin’s launch, Nubank identified a clear opportunity to accelerate their refactor at a fraction of the engineering hours. Migration or large refactoring tasks are often fantastic projects for Devin: after investing a small, fixed cost to teach Devin how to approach sub-tasks, Devin can go and complete the migration autonomously. A human is kept in the loop just to manage the project and approve Devin’s changes.

The Solution: Custom ETL Migration Devin

A task of this magnitude, with the vast number of variations that it had, was a ripe opportunity for fine-tuning. The Nubank team helped to collect examples of previous migrations their engineers had done manually, some of which were fed to Devin for fine-tuning. The rest were used to create a benchmark evaluation set. Against this evaluation set, we observed a doubling of Devin’s task completion scores after fine-tuning, as well as a 4x improvement in task speed. Roughly 40 minutes per sub-task dropped to 10, which made the whole migration start to look much cheaper and less time-consuming, allowing the company to devote more energy to new business and new value creation instead.

Devin contributed to its own speed improvements by building itself classical tools and scripts it would later use on the most common, mechanical components of the migration. For instance, detecting the country extension of a data class (either ‘br’, ‘co’, or ‘mx’) based on its file path was a few-step process for each sub-task. Devin’s script automatically turned this into a single step executable — improvements from which added up immensely across all tens of thousands of sub-tasks.

There is also a compounding advantage on Devin’s learning. In the first weeks, it was common to see outstanding errors to fix, or small things Devin wasn’t sure how to solve. But as Devin saw more examples and gained familiarity with the task, it started to avoid rabbit holes more often and find faster solutions to previously-seen errors and edge cases. Much like a human engineer, we observed obvious speed and reliability improvements with every day Devin worked on the migration.

Results: Delivering an 8-12x faster migration, lifting a burden from every engineer, and slashing migration costs by 20x.

“Devin provided an easy way to reduce the number of engineering hours for the migration, in a way that was more stable and less prone to human error. Rather than engineers having to work across several files and complete an entire migration task 100%, they could just review Devin’s changes, make minor adjustments, then merge their PR”

Jose Carlos Castro, Senior Product Manager

8-12x efficiency gains This is calculated by comparing the typical engineering hours required to complete a data class migration task against the total engineering hours spent prompting and reviewing Devin’s work on the same task.
Over 20x cost savings on scope of the migration delegated to Devin This is calculated by comparing the cost of running Devin versus the hourly cost of an engineer completing that task. The significant savings are heavily driven by speed of task execution and cost effectiveness of Devin relative to human engineering time – it does not even consider the value captured by completing the entire project months ahead of schedule!
Fewer dreaded migration tasks for Nubank engineers

Choose a plan
that's right for you

Core
Pay as you go
Get started
Includes:
Key capabilities:
  • Autonomous task completion
  • Devin IDE
  • Devin Search
  • Devin Wiki
  • Learns over time
Usage:
  • Unlimited users
  • Share and collaborate
  • Up to 10 concurrent Devin sessions
  • No monthly commitment, pay-as-you-go plan
  • Auto-reload settings for on-demand consumption
Team
$500/month
Get started
Everything in Core, plus:
Key capabilities:
  • Devin API
  • Access to early feature releases and research previews
Usage:
  • Unlimited concurrent sessions
  • 250 ACUs included monthly Devin's unit of work is an Agent Compute Unit, or ACU. It's a normalized measure of the resources used by Devin.
Account Support:
  • Dedicated Slack Connect channel for support
  • Optional onboarding call with the Cognition team
Enterprise
Custom pricing
Contact us
Everything in Team, plus:
Key capabilities:
  • Access to Devin Enterprise Devin Enterprise is the most capable version of Devin, available for users on the Enterprise plan.
  • Access to Custom Devins Custom Devins are fine-tuned versions of Devin specialized for specific use cases and/or proprietary datasets.
Security:
  • Deploy in your virtual private cloud (VPC)
  • SAML/OIDC SSO
  • Centralized enterprise admin controls
  • Teamspace isolation
Account Support:
  • Dedicated account team
  • Custom terms
  • Centralized billing and usage analytics across multiple Devin organizations

Compare plans

Core

Pay as you go

Get started
Team

$500/month

Get started
Enterprise

Custom pricing

Contact us
Core
Team
Enterprise
Pay as you go
Get started
$500/month
Get started
Custom pricing
Contact us
Key capabilities:
  • Access to Devin 
    • Powered by Devin 
    • Powered by Devin 
    • Powered by Devin Enterprise  Devin Enterprise is the most capable version of Devin, available for users on the Enterprise plan.
  • Autonomous task completion 
  • Learns over time 
  • Devin IDE 
  • Devin Search 
  • Devin Wiki 
  • Devin API 
Integrations:
  • Slack 
  • GitHub 
  • Custom git provider support 
Usage
  • Monthly ACUs  Devin's unit of work is an Agent Compute Unit, or ACU. It's a normalized measure of the resources used by Devin.
    • Pay-as-you-go with a starting minimum of $20 ($2.25 per ACU) 
    • 250 ($2.00 per ACU) 
    • Custom 
  • Auto-reload ACUs 
    • Custom 
  • Seats 
    • Unlimited 
    • Unlimited 
    • Unlimited 
  • Share and collaborate 
  • Concurrent sessions 
    • Up to 10 concurrent sessions 
    • Unlimited 
    • Unlimited 
Account Support:
  • Dedicated Slack Connect channel for support 
  • Optional onboarding call with the Cognition team 
    • Dedicated account and engineering support 
  • Access to early feature releases and research previews 
    • Custom 
  • Dedicated account team 
  • Custom terms 
  • Enterprise accounts  Enterprise accounts enable centralized management of multiple Devin organizations
Security:
  • Deploy in your virtual private cloud (VPC) 
  • SAML/OIDC SSO 
  • Centralized enterprise admin controls 
  • Teamspace isolation 

FAQs

Contact us at support@cognition.ai
to learn more

  • What is an ACU?

    Devin's unit of work is an Agent Compute Unit, or ACU. It's a normalized measure of the computing resources Devin uses to complete a task, such as virtual machine time, model inference, and networking bandwidth.

  • How am I charged for ACUs?

    Depending on your plan, you'll have different options to purchase ACUs. With the Core plan, you only pay for what you use (at $2.25 per ACU) and can customize your account's auto-recharge settings to make sure your team is never blocked.

    With the Teams plan, your subscription will include 250 ACUs for $2.00 per ACU. You can also set your account's auto-recharge settings to continue consuming ACUs on-demand after you've used up the credits included in your subscription.

  • How are ACUs consumed?

    Devin will only consume ACUs when actively working on a task or when the virtual machine is running. There are many variables that impact ACU consumption including task complexity, prompt quality (or specificity), size of the codebase or number of files Devin touches/edits, and session runtime.

    For more information about ACUs and our billing model please visit the docs.

  • Do I get charged more for using the API?

    If you're on a Teams or Enterprise plan and have access to the Devin API, you'll only consume ACUs based on the sessions you start programmatically via the API. There are no additional costs to using the API.

  • Who owns the code generated by Devin?

    You do! Regardless of the plan, all inputs and outputs are considered your intellectual property. For more information, we recommend reviewing the Terms of Service for each plan.

  • Additional questions?

    Contact us at support@cognition.ai to learn more.