# Noah Labs Boosts AI Efficiency with Google Cloud's Gemma 4 Model

> By leveraging Google Cloud's Gemma 4 model, Noah Labs has revolutionized code modernization, delivering unprecedented accuracy and efficiency in legacy systems. Discover how their innovative Sentinel tool is set to redefine software development practices.

**Source**: cloud.google.com | **Published**: 2026-11-18 | **Type**: case_study

## Key Facts

- Noah Labs achieves 3x higher accuracy per dollar, enhancing cost-effectiveness in AI deployment.
- Gemma 4 31B model delivers 65.6% faster response speeds, boosting developer productivity significantly.
- Evaluation spend reduced by 3.1x to 7.5x, indicating strong financial efficiency in AI model selection.
- 97.6% coding accuracy positions Noah Labs as a leader, revealing competitive edge in software modernization.
- Lean hardware configurations outperform larger models, signaling a strategic shift in AI deployment practices.

## Summary

\## Summary
Noah Labs, a technology company based in Palo Alto, faced challenges in modernizing legacy code for its clients, including defense contractors and banks. By deploying the Gemma 4 model on Google Cloud, they achieved a 3x higher accuracy per dollar in code modernization and significantly reduced evaluation costs, resulting in a 97.6% coding accuracy across tested models.

\## Background
Noah Labs specializes in developing AI tools for mission-critical software and autonomous systems. The company serves various sectors, including defense, government, and finance. Before implementing their AI solution, Noah Labs struggled with the inefficiencies of modernizing complex, aging codebases, which required fast and reliable AI systems to support developers in their workflows.

\## Challenge
The primary challenge was to enhance the efficiency of code refactoring processes while maintaining high accuracy and minimizing costs. Traditional approaches relied on larger AI models, which often resulted in slower response times and higher infrastructure expenses, disrupting developer workflows.

\## Solution
Noah Labs developed Sentinel, a structured orchestration framework that integrates the Gemma 4 model. This system manages five critical stages of the engineering workflow: context retrieval, invoking development tools, orchestrating tasks, running automated verification, and conducting structured reviews. By surrounding the AI model with proper context and tools, Sentinel ensures that code modifications are thoroughly checked before human review, enhancing productivity.

\## Results
The deployment of Sentinel on Google Cloud led to a 3.1x to 7.5x reduction in evaluation spending compared to larger models. The system achieved 97.6% coding accuracy and delivered 65.6% faster response speeds for developer workflows. Noah Labs reported 6.082 accuracy points per dollar, significantly outperforming larger models like GPT-OSS-120B and Granite-4.1 in both cost and precision.

\## Key Insights
Focusing on completed, validated work rather than model size can lead to better engineering outcomes. Efficient orchestration of AI models within a structured framework enhances productivity and reduces costs, allowing for practical AI deployment in real-world applications.

\## Customer Testimonial
"Noah Labs is advancing practical AI for legacy software modernization, helping engineering teams turn complex codebases into understandable, testable, and maintainable systems," says Murat Can Isik, Co-founder and CEO at Noah Labs.

## Entities

- **Companies**: Noah Labs, Google Cloud, Google DeepMind
- **Products**: Gemma 4, Gemini, Compute Engine
- **People**: Murat Can Isik, Berk Yilmaz

## Key Concepts

AI tools for software modernization, Code refactoring, Engineering workflows, Cost efficiency in AI, Performance evaluation of AI models, Legacy code management, Automated verification, Cloud deployment

## Definitions

- **Code modernization**: The process of updating legacy codebases to improve maintainability and performance using modern technologies.
- **Sentinel**: A structured orchestration framework developed by Noah Labs to enhance AI-driven code refactoring and verification.
- **Gemma 4**: A dense AI model from Google DeepMind that powers Noah Labs' Sentinel system, offering high accuracy and efficiency.
- **Automated verification**: The use of AI systems to automatically check code changes for correctness before human review.
- **Continuous iteration cycles**: A development process where code is frequently modified, tested, and refined in rapid succession.

## Use Cases

- Modernizing legacy codebases for defense contractors
- Supporting government agencies with AI tools
- Enhancing software development workflows for banks
- Improving navigation reliability in autonomous systems
- Reducing operational costs for software engineering teams
- Facilitating faster code testing and validation

## Frequently Asked Questions

**What benefits does Noah Labs provide for software teams?**

Noah Labs offers AI tools that significantly enhance the accuracy and speed of code modernization, allowing teams to refactor legacy code efficiently. Their solutions help maintain high engineering quality while reducing operational costs.

**How does Sentinel improve the code refactoring process?**

Sentinel orchestrates the engineering workflow by managing context retrieval, invoking development tools, and running automated verification checks. This structured approach ensures that code modifications are thoroughly validated before human review.

**What is the significance of using Gemma 4 in Noah Labs' solutions?**

Gemma 4 is a high-performance AI model that delivers exceptional accuracy and speed, making it ideal for powering Sentinel. Its efficiency allows software teams to iterate quickly without compromising on quality.

**How does Google Cloud support Noah Labs' operations?**

Google Cloud provides the infrastructure necessary for Noah Labs to deploy its AI solutions effectively. It enables cost-efficient operations and supports the scalability required for modern software development.

**What are the future plans for Noah Labs?**

Noah Labs aims to develop deeper automated verification features to further streamline the code assessment process. They are also exploring additional deployment options using Gemini models to enhance their offerings.

## Links

- [Read on Welcome.AI](https://welcome.ai/content/noah-labs-boosts-ai-efficiency-with-google-clouds-gemma-4-model)
- [Original source](https://cloud.google.com/customers/noah-labs)
- [Noah Labs](https://welcome.ai/company/noah-labs): Featured company

---

Source: Welcome.AI | https://welcome.ai/content/noah-labs-boosts-ai-efficiency-with-google-clouds-gemma-4-model