Connecting...

Client

The client is a leading Insurance Group with controlled operations across Australia and New Zealand, underwriting over $14 billion in premium per annum through a diverse portfolio of established insurance brands. The organisation operates across three core business divisions aligned to specific customer needs: Direct Insurance Australia, Intermediated Insurance Australia, and General Insurance New Zealand.

Challenge

Operating across a vast digital landscape, the client faced significant operational friction and bottlenecks within its digital and product engineering workflows. Prior to implementing a modernised solution, development teams relied entirely on manual Postman configurations for making Application Programming Interface (API) calls. This reliance on manual testing harnesses created severe operational challenges across the software development lifecycle:

  • Repetitive Manual Overhead: Engineering teams were forced to construct, edit, and troubleshoot Postman API requests by hand for every individual testing scenario, consuming valuable development hours.

  • Multi-Environment Friction: Operations and test execution had to be repeated manually across multiple isolated deployment environments, as no automated framework existed to streamline cross-environment validation.

  • Inefficient Content Management: Content updates sourced from internal repositories like Confluence required manual application across disparate communication templates. Validating these changes necessitated painstaking, environment-by-environment manual searches.

  • Elevated Risk Profile: Because payload updates, environment switching, and data cross-referencing relied exclusively on human accuracy rather than a standardised, automated workflow, the entire process was highly error-prone, time-consuming, and difficult to audit.

Solution

To eliminate the operational friction associated with manual API testing and fragmented test harnesses, FinXL designed and built an intelligent SmartCOMM AI Agent. The solution unified platform analysis, automated testing, environment validation, and governance into a single, cohesive workflow.

Key delivery and engineering highlights included:

  • Native IDE Integration: The AI Agent was integrated directly into Visual Studio Code via GitHub Copilot Chat. This native setup allowed developers to interact with the capability within their primary integrated development environment (IDE), completely eliminating context-switching and avoiding the need to install standalone software.

  • Natural Language Operational Control: Leveraging the Model Context Protocol (MCP), Copilot was connected directly to SmartCOMM’s OAuth1-secured REST APIs. This enabled software engineers to trigger and execute over 120 platform operations using simple natural language prompts.

  • Automated Core Workflows: The agent fully automated unit test generation, user and environment validations, and platform governance checks. This removed the requirement for developers to build manual test scenarios or manually cross-reference Confluence documentation.

  • Enterprise AI Governance Approval: FinXL worked closely with the client’s internal AI Governance function to conduct thorough risk and compliance reviews, securing formal architecture approval to ensure a safe, structured rollout.

Business Value

By engaging FinXL IT Professional Services to deploy the custom SmartCOMM AI Agent, the organisation unlocked significant strategic advantages and operational efficiencies:

  • Eliminated Onboarding Friction: Because the AI Agent operates seamlessly within the developers' existing daily toolset (Visual Studio Code), onboarding requires zero setup time or additional software overhead.

  • Drastic Efficiency and Time Savings: Automating unit test generation, multi-environment validation, and governance checks significantly reduced the hours previously wasted on manual Postman scripting and manual cross-checks.

  • Scalable, Community-Driven Capability: Designed as a "living" capability, the AI Agent allows internal engineering teams to build, share, and contribute new skills via Git, ensuring the tool continually evolves alongside real-world development needs.

  • De-Risked AI Adoption: Successfully passing rigorous internal AI Governance reviews ensured the solution met strict enterprise compliance, security, and data handling standards, establishing a clear pathway for safe AI integration across broader business units.

Accelerate Your Digital & Product Engineering

Is your organisation struggling with manual testing friction or developer velocity bottlenecks? FinXL can help you design, govern, and deploy intelligent AI agents and automated workflows to boost delivery performance.

Contact FinXL Today

See More