Enhancing the software development lifecycle with AI
FINANCIAL SERVICES
Enhancing the software development lifecycle with AI
At a glance
An international financial services organisation sought to modernise its platforms and embed AI into the current software development lifecycle (SDLC). Specific targets included a 30% to 40% efficiency uptick without increasing internal headcount.
Confronted with a time-critical integration deadline, the enterprise needed to increase engineering capacity and accelerate BAU delivery, all whilst maintaining sector-compliant coding standards, security, and traceability.
Any solution would need to be scalable and minimise adding to the company’s internal headcount.
FDM rapidly assigned an Agile delivery team of AI-enabled consultants aligned to the company’s existing delivery model, with a focus on reducing time-on-task for targeted activities.
FDM Practices
- Software Engineering
Industry
Financial Services
Tech stack
React
.NET
Azure DevOps
Azure Cloud
GitHub Copilot
Claude
Figma
Impact
70% – 80%
of coding tasks automated
~1 hour
to complete coding tasks, down from three days
4x
faster processing for API refactoring
These productivity gains, achieved by integrating new talent and AI-driven tools into the engineering process, means that the client is well-placed to reach their ambitious 30% to 40% efficiency goals.
Role-ready talent embeds AI from day one
Following a four-week custom training programme in the FDM Skills Lab, we assigned FDM Consultants aligned to the client’s objectives: junior AI-enabled developers trained in business-critical tools and AI products, working under light-touch senior supervision.
The FDM Consultants formed two teams, with a Scrum Master and two Lead Engineers guiding four junior developers each, supported by a Business Analyst.
The junior consultants used AI agent-enhanced Integrated Development Environments (IDEs) for “first-pass” code creation, performing tasks like writing boilerplate code, generating unit tests and documentation, or summarising complex code blocks to speed up understanding. Seniors then focused on the complex architecture and fine-tuning the AI-generated output.
Integrating AI with robust guardrails
The FDM Consultants were divided into parallel workstreams to best address the client’s challenges efficiently.
Workstream A focused on rapid completion of time-sensitive deliverables, including:
- Integrating an acquired SaaS platform into the client’s customer portals, using React front-end and .NET back-end
- Implementing secure file upload/download, dynamic dashboards, and data integration
Workstream B tackled a research and development project to build a governed prompt library. This would act as a repository of standardised AI prompts capturing the client’s architectural patterns, coding standards, APIs, and approved sources.
The teams also embedded rigorous and robust pipeline guardrails within Azure DevOps CI/CD to oversee all AI-assisted contributions across a range of functions, including branch policies, automated test suites, dependency checks, and static code scans.
These security measures allowed for GitHub Copilot and Claude to be utilised for rapid code drafting, documentation, and test generation, whilst ensuring a human oversight for critical code review and all security-sensitive actions.
“The speed at which the FDM Consultants integrated was invaluable. Through their collaboration with our Project Team, they created a highly productive environment. I was especially impressed with their attention to detail when managing error states and form fields, which ensured a smooth process and greatly contributed to our success.”
Client’s Head of Transformation
SDLC reinvigorated with AI-driven engineering
The FDM Consultants’ work on the client’s SDLC delivered several significant outcomes, including:
Our client integrated our consultants’ outputs into ongoing pipelines and continued to apply prompt-driven delivery more widely. It also planned further innovation, including testing how teams could move from requirements to a working application within 24 hours, alongside expanding UI automation and using AI to modernise legacy services.
“Our AI-driven pipeline has significantly accelerated development timelines. For example: the generation of a complete frontend user journey in just three days, whereas a traditional development team required over two weeks to deliver a comparable outcome.”
Jude Craig, Software Engineer, FDM Consultant
Conclusion
Through the rapid assignment of AI-enabled FDM Consultants, we supported modernisation whilst protecting standards, security, and traceability across the software development lifecycle.
Continuing these successes, the client will build upon the foundations laid by the FDM Consultants, expanding the prompt library, scaling AI enablement, formalising tooling against client constraints, and establishing an AI SDLC scorecard for ongoing measurement. The client is now using AI tools to reshape their SDLC, including forward and reserve engineering, plus security and governance.
