This document outlines the comprehensive development roadmap for our Co-Pilot Onixia system, designed to enhance IT support operations across Campus, Branch, and Network Monitoring teams and eventually the entire estate. The roadmap is structured in six progressive phases, from initial planning through to full deployment, with clear status indicators for each stage. This strategic implementation plan details key integration requirements, access enablement processes, testing protocols, and rollout strategies to ensure a successful transformation of our service desk operations through AI-assisted support capabilities.
Status: ✅ Completed
The initial phase of our Co-Pilot Agent development has been successfully completed, establishing the foundational elements necessary for the project's success. This critical planning stage involved comprehensive analysis of current service desk operations and identification of opportunities for AI-assisted enhancement across all support domains.
We've carefully defined distinct use cases for each of our three primary support areas: Campus, Branch, and Network Monitoring teams. Each domain presents unique challenges and opportunities for AI-assisted support, requiring tailored approaches to maximise effectiveness. These use cases were developed through collaborative workshops with team leaders to ensure alignment with actual operational needs.
Clear objectives have been established to guide the development process, focusing on three primary outcomes: implementation of guided troubleshooting workflows to improve first-call resolution rates, automation of logging processes to reduce administrative burden, and intelligent escalation protocols to ensure issues reach appropriate specialists promptly.
We've identified the critical systems requiring integration with the Co-Pilot Agent, primarily our IT Service Management (ITSM) platform and Enterprise Data Warehouse (EDW). These integrations will enable seamless incident management workflows and comprehensive reporting capabilities within the agent interface.
The successful completion of Phase 1 has provided a solid foundation for subsequent development activities, with clearly defined requirements and objectives guiding our technical approach. This thorough planning effort ensures that development resources are appropriately allocated to deliver maximum value to our service desk operations.
Status: 🔄 In Progress
The second phase of our roadmap focuses on the core development activities required to create a functional Co-Pilot Agent with robust integration capabilities. This phase represents the primary technical implementation effort, building upon the requirements established during planning.
Our development team is actively constructing guided troubleshooting flows tailored to each Service Desk domain. These decision-tree structures will guide agents through step-by-step diagnostic procedures, ensuring consistent methodology regardless of agent experience level. Each flow incorporates domain-specific technical knowledge and best practices gathered from our most experienced support specialists.
The flows are being developed with flexibility in mind, allowing for continuous refinement based on emerging issues and evolving infrastructure. Initial focus has been placed on high-volume incident categories to maximise immediate impact upon deployment.
Parallel to flow development, we are defining the technical integration architecture that will enable seamless interaction between the Co-Pilot Agent and our core IT management systems. This integration framework encompasses:
The final component of our current development effort involves creating the query logic that will eventually power the Co-Pilot's reporting capabilities. Working closely with our data architecture team, we are designing optimised query structures that will enable agents to access relevant historical data without requiring specialised SQL knowledge. These queries will support trend analysis, comparative incident metrics, and performance dashboards directly within the agent interface.
Phase 2 represents the most technically intensive portion of our roadmap, with multiple workstreams progressing concurrently. Regular integration testing is being conducted to ensure compatibility between newly developed components, with an emphasis on maintaining alignment with the requirements established during Phase 1.
Status: 🔄 In Progress
Phase 3 focuses on establishing the necessary technical environments and system access required to support Co-Pilot Agent development, testing, and eventual production deployment. This infrastructure-focused phase runs parallel to development activities and is critical for enabling proper integration testing.
Access to our ITSM development environment is being coordinated by Melissa Oosthuisen, who serves as the primary liaison with the ITSM platform team. This segregated environment provides a safe space for testing Co-Pilot integration without impacting production operations. It replicates all essential ITSM functionality including incident management, knowledge base access, and workflow automation capabilities.
The development environment includes sample incident data that accurately reflects our production patterns, enabling realistic testing scenarios during integration validation.
We are establishing secure API connectivity between the Co-Pilot Agent and our ITSM platform to enable programmatic ticket handling. This includes configuration of authentication protocols, endpoint mapping, and permission scoping to ensure appropriate access controls. The API integration will support real-time incident creation, updates, and status tracking without requiring direct user intervention.
Our security team is closely involved in this process to ensure all connections adhere to our data protection standards and encryption requirements.
Engagement with Linda from our data management team has been initiated to secure access to relevant EDW datasets. This access will enable the Co-Pilot to generate meaningful analytics and reports directly within the agent interface. We are working to establish appropriate read-only permissions that provide sufficient data access while maintaining our data governance principles.
This integration will allow for dynamic reporting capabilities that transform raw incident data into actionable insights for both agents and management.
The environment and access enablement phase represents a critical dependency for subsequent testing activities. Our infrastructure team is prioritising these requests to prevent any delays in the overall project timeline. Regular status updates are being provided to ensure visibility into the progress of each access request, with escalation protocols in place for any items at risk of delay.
Status: ⏳ Upcoming
Following the completion of development and access enablement phases, we will initiate comprehensive testing of the Co-Pilot Agent to ensure it meets both functional requirements and quality standards. This structured QA process will validate all aspects of the system before proceeding to pilot deployment.
Our testing strategy will employ a multi-layered approach to ensure thorough validation of all Co-Pilot components and integrations:
Individual components will be tested in isolation to verify correct functioning of each troubleshooting flow, integration point, and reporting query. This will include validation of decision logic, data transformations, and error handling capabilities.
End-to-end tests will validate the complete workflow from initial incident capture through to resolution, including all touchpoints with ITSM and EDW systems. Special attention will be paid to data consistency across system boundaries.
Selected service desk agents will perform guided testing of real-world scenarios to validate usability and effectiveness from an end-user perspective. This will ensure the Co-Pilot enhances rather than complicates the support process.

All identified issues will be captured in our defect tracking system with appropriate severity classifications. Critical defects that impact core functionality will be addressed immediately, while minor issues may be prioritised for post-pilot resolution based on impact assessment. Regular defect review meetings will ensure appropriate visibility and prioritisation of required fixes.
The testing phase will conclude with a formal Go/No-Go decision point, requiring sign-off from key stakeholders before proceeding to pilot deployment. This checkpoint ensures that only a quality-verified solution proceeds to user-facing implementation, minimising risk to service desk operations.
Status: ⏳ Upcoming
The pilot deployment represents our first controlled introduction of the Co-Pilot Agent to actual service desk operations. This phase will provide valuable real-world validation while limiting potential operational impact to a manageable subset of our support environment.
Our pilot approach will follow a carefully structured implementation plan designed to maximise learning while minimising risk:
The pilot will be evaluated against several key performance indicators to determine readiness for full deployment:
Target improvement in average time to resolution for incidents handled with Co-Pilot assistance
Percentage of incidents correctly categorised and routed to appropriate support teams
Percentage of pilot users reporting positive impact on their support capabilities
Throughout the pilot, we will collect detailed feedback from participating agents through structured surveys and informal discussion sessions. This qualitative feedback will complement our quantitative metrics to provide a comprehensive view of the Co-Pilot's effectiveness. Any identified issues or enhancement opportunities will be documented and prioritised for resolution before proceeding to full deployment.
The pilot phase will conclude with a formal assessment report summarising all findings and providing a clear recommendation regarding readiness for organisation-wide implementation.
Status: ⏳ Upcoming
The final phase of our Co-Pilot Agent roadmap involves comprehensive deployment across all service desk operations and establishment of ongoing monitoring processes to ensure continued effectiveness and improvement.
Following successful pilot validation, we will implement a phased rollout strategy to bring the Co-Pilot Agent to all service desk teams:
Each deployment phase will include comprehensive training sessions, readily available documentation, and dedicated floor support during the initial transition period. We will maintain heightened technical support availability throughout the deployment process to address any emerging issues promptly.

Once fully deployed, we will implement ongoing performance monitoring through several mechanisms:
To ensure the Co-Pilot Agent remains effective and continues to evolve with our service desk operations, we will establish a structured continuous improvement process:
Ongoing gathering of performance metrics, user feedback, and operational insights to identify potential enhancement opportunities
Regular assessment of collected data to identify high-impact improvement areas and prioritise development efforts
Agile implementation of prioritised improvements through regular development sprints
Controlled release of enhancements with appropriate testing and user validation
This iterative improvement process will ensure the Co-Pilot Agent evolves to meet changing operational needs and continues to deliver increasing value over time. We will maintain a public enhancement roadmap to provide visibility into upcoming improvements and allow stakeholders to contribute suggestions for future development priorities.
By establishing robust deployment, monitoring, and improvement processes, we will ensure the long-term success and sustainability of the Co-Pilot Agent as a core component of our service desk operations.
Co-Pilot Agent Development Roadmap