Customers

Lemonade Enhances Developer Productivity and Navigates Complex Codebase with Augment

Facing challenges with a 10-million-line monolithic codebase in Ruby, Lemonade used Augment to help their TypeScript-focused development team navigate and understand legacy code, resulting in significant productivity gains and faster onboarding for new developers.

The Challenge

Lemonade, a technology-driven insurance company, operates with a vast and intricate codebase. Their monolithic repository contains nearly 10 million lines of code, much of which is undocumented and written in Ruby—a language unfamiliar to the majority of their current developers, who primarily use TypeScript.

Navigating this extensive codebase was a significant hurdle, especially for new team members. "It's really hard to find yourself in the monolith," said Chen Kinnrot, Engineering Manager at Lemonade. "When you don't speak the language, it's really hard to find yourself."

Why Augment:

Seeking to enhance developer productivity and code comprehension, Lemonade explored AI-based coding tools and chose Augment for several key reasons:

  • Deep Context Awareness: Augment's ability to understand and process large codebases stood out. It provided detailed answers on complex areas, even in Ruby. "It pinpoints you to the right line of code or area that you need to look at, saving days of effort."
  • Unmatched Inline Completions: Augment's inline completion feature allowed developers to edit code in the middle of a line with automatic suggestions. "One of the most killer features was the inline completion... I didn't see any tool that can suggest you changes automatically in that area, and Augment can."
  • Scalability and Adoption: Augment’s Slackbot and chat features enhanced engagement and usability across Lemonade’s teams. "We saw a huge ramp-up in active developers once we released the Slackbot."

Results and Benefits:

Implementing Augment led to significant improvements for Lemonade's engineering team:

  • Increased Developer Productivity: Augment's speed and accuracy helped engineers complete tasks more efficiently. Developers consistently noted that it outperformed Copilot in both speed and quality.
  • Improved Code Navigation and Understanding: The tool empowered developers to quickly locate relevant sections of the monolith, even in unfamiliar codebases, significantly reducing time spent on code exploration. "It helps new developers find themselves in the code, especially when they have tasks that touch the monolith."
  • Accelerated Onboarding for New Developers: Augment proved instrumental in onboarding engineers to the legacy codebase, particularly those unfamiliar with Ruby.
    "It helps new developers navigate the code with ease. Without Augment, it would take days to understand certain sections."
  • Better Team Collaboration: By providing actionable insights and explanations, Augment reduced dependencies between developers and other teams, allowing engineers to focus on critical tasks.
    "Product managers and engineers alike find value in Augment—it bridges gaps across roles."
  • Future-Proofing Development: Lemonade optimized workflows, preparing for a future where developers take on more design and architectural responsibilities while relying on AI for routine coding tasks. "If we keep this acceleration, developers will focus on design and architecture, letting AI handle the simpler tasks.".

About the Company

Powered by AI and social impact, Lemonade is a purpose-built, technology-first insurance carrier.As a customer-centric tech company, Lemonade creates an insurance experience across Renters, Home, Pet, Car, and Life that is smart, instant, and delightful. Their team of 1,200+ Lemonade Makers make it possible for over 2M customers throughout the US, UK and Europe to get coverage instantly, with nearly half of claims paid in a matter of seconds.

Industry: Insurance Technology, Artificial Intelligence, Social Impact

Location: New York

Team Size: 226 Engineers

Languages, Frameworks, Libraries: Typescript, React, Rust, Python, Ruby

IDE: VS Code, IntelliJ

Molisha Shah

GTM and Customer Champion

Keep reading with us.
Scott Dietzen
February 7, 2025
The rise of specialty models: 6 predictions for AI in 2025
Molisha Shah
December 18, 2024
Beyond code generation: What if your AI actually understood your codebase?
Chris Kelly
February 14, 2025
Augment <3 Vim