Linear infrastructure engineering
Generative AI adoption in software engineering with Claude
A structured approach to moving from isolated AI experiments to AI-assisted software development workflows using Claude.
Industry
Software solutions for large-scale infrastructure projects
Challenge
Integrate AI to transform and accelerate software delivery cycles, moving from SDLC to AI-DLC
Impact
A validated roadmap to reduce development cycles by 60% in under three months
Expertise
LLM benchmarking
AI pair programming
Cross-functional engineering
DevOps
Adoption planning
Technical knowledge-sharing sessions
Tech Stack
Claude
Java/Spring
React
PostgreSQL
Kubernetes
Jenkins
The software company develops specialised, tailored solutions for managing large-scale infrastructure projects.
Since 2021, our ongoing involvement with its technology teams has given us a deep understanding of its development requirements. Working as an integrated engineering team, we supported the company in transforming its software engineering practices.
The goal was to establish a reliable foundation for generative AI-assisted development. This meant moving beyond isolated experiments towards AI-driven development workflows (AI-DLC).
Within three months, the transition had already delivered measurable improvements in development speed and team collaboration.
Key Challenges
AI tools that can generate new software features are changing how quickly companies can develop products. To remain competitive, the company needed to rethink its development processes.
- Faster development: Reduce delivery times and accelerate proofs of concept (PoCs). These early tests are essential for validating software architecture before large-scale development.
- Protecting existing systems: Integrate large language models (LLMs) without compromising industry expertise, security or the stability of the existing codebase.
- Changing development practices: Move from the traditional Software Development Life Cycle (SDLC) to an AI-assisted approach. This required a clear adoption plan for all engineering teams.
Collaborative Execution
As the company's technology and AI partner, we introduced an approach based on short development cycles and regular deliverables.
This helped technology, data and product teams stay aligned throughout the project.Following an evaluation of several LLMs, we selected Claude for its analytical capabilities. Our AI deployment engineers then organised the implementation into three stages.
First, they configured dedicated environments to handle complex algorithmic queries using Claude.Next, they ran AI pair programming sessions to review the existing codebase and identify areas for improvement.
Finally, drawing on our expertise as a certified Anthropic Claude Partner Network member, we introduced weekly knowledge-sharing sessions. These sessions allowed engineers to review real use cases, exchange feedback and establish consistent ways of using the tool.
Measurable Impact
The new operating model has already improved engineering productivity.
- Faster development cycles: Automating complex technical tasks has given engineers more time to explore ideas and improve their work. These gains are already visible in PoC timelines, supporting a roadmap to reduce the most demanding delivery cycles from eight months to three.
- Broader engineering skills: Targeted AI support has helped engineers work across traditional technical boundaries. Backend engineers can now independently take on frontend development tasks using React. This flexibility has created a more versatile team, making it easier to redistribute work and respond to new client requirements.
- A shift in engineering culture: Adoption of the AI-DLC model continues to progress. Regular knowledge-sharing sessions help teams align their development practices and establish consistent working methods across the technology department.
Discover more case studies
Every project is unique, but our commitment remains the same: to co-deliver, de-risk, and move products forward—regardless of context or constraints.
