Roadmap Alignment

## Summary ### Operational Challenges and Team Efficiency The meeting opened with a discussion about the negative impact of back-to-back meetings on productivity, noting that constant context switching often prevents meaningful work from being completed and leads to suggested actions not being taken. A strategy was proposed to mitigate this by blocking out large, uninterrupted time slots-specifically, three-hour blocks-for focused work, with a preference for scheduling meetings in the morning to free up afternoons for deep work. ### Technical Tasks and Process Documentation A key action item involves the **NG downloader and image-to-PDF convergence**, which is needed for onboarding a client's data. While the task is urgent, there is a significant knowledge gap as no one on the team is currently familiar with the process. The decision was made to proceed with the work but to ensure it is properly documented as a standard onboarding exercise. This documentation will cover the entire workflow, including the location of scripts, the file formats they handle, and the upstream and downstream processes, to prevent similar knowledge silos in the future. Concurrently, there was an unresolved issue regarding **access to an Azure OpenAI instance**. Access requests had been submitted but were seemingly stalled in the system. A workaround was successfully executed to grant access, but the underlying problem with the request system itself remains and requires further investigation with the service desk. ### Roadmap and Priority Planning for Q4 The team reviewed and adjusted the roadmap for November and December, introducing new priorities that have emerged. **November Priorities:** - **Invoice Processing Scaling:** The system is now processing close to 3,500 invoices per day, exceeding initial goals. The immediate focus is shifting from pure scaling to **comprehensive tracking and dashboard creation**. This is crucial for monitoring API usage (like OpenAI token consumption) and providing data-driven evidence for any future capacity discussions with vendors. - **DSS Logic Fixes:** A dedicated effort is required to address underlying logic issues within the Data Services (DSS) and UBM systems. This has been deprioritized previously but is now being elevated to ensure system stability. - **"Unified Report" Development:** This initiative aims to create a comprehensive dashboard that provides a single view of a customer's status, from onboarding through the entire builder stack. The approach will be to have stakeholders provide specific Excel examples of their desired reports to streamline requirements gathering and set realistic expectations. **December and Beyond:** The team expressed concern over an overly ambitious December schedule. The strategy is to push for clearly defined requirements from stakeholders upfront and to resist becoming a de facto research and reporting team for other departments. The goal is to transition from reactive firefighting to a more stable operational state, which will enable work on strategic projects in the new year. One such project for the following year is a **reporting API/database**, identified as a critical feature for potential customers. ### Team Structure and Future Initiatives A significant topic was the future structure of the team, particularly a planned transition where a key member would move into a Bill Pay-focused role. This transition is contingent upon achieving stability in the Data Services domain and is not considered feasible until the current operational challenges are resolved. The long-term vision is for the UBM team to evolve into a proper platform team, freed from daily QA and data validation issues, so it can focus on building new features like the reporting API. A plan was set to begin detailed planning for 2026 initiatives in a shared document to further refine timelines and responsibilities. ### Arcadia Client Migration Finally, the upcoming migration for the Arcadia client was discussed. Two technical approaches for data ingestion are being considered: treating Arcadia similarly to an existing client (Biowatcher) or developing a new pipeline based on their API structure. A proof-of-concept will be initiated to determine the best path forward. It was strongly emphasized that the team's responsibility is limited to building the pipeline, with ongoing maintenance and error resolution remaining the responsibility of the Arcadia Customer Success Managers.