This pack is for the Operations, Treasury & Business Management team only.
Enter your team code to unlock.
Aurora Nexus Global's operations and treasury functions form the backbone of its financial controls and balance sheet management. Group Operations manages the collections and servicing infrastructure across ANG's retail and SME loan portfolios. Group Financial Markets oversees Asset and Liability Management and treasury positioning. Group Finance consolidates the full picture for EXCO. Together, these three teams are responsible for the accuracy, timeliness, and reliability of the financial intelligence that drives board-level decisions. Right now, they are producing that intelligence by hand every Monday morning, from three systems that do not talk to each other, and the numbers do not always agree.
Every Monday morning, three separate teams begin the same task: pulling data from different systems to build the weekly performance reports that go to EXCO. Group Operations extracts collections data from the loan management system, the arrears management platform, and a legacy system that was never formally decommissioned. Group Financial Markets pulls ALM positions from the treasury management system and reconciles them manually against the funding ledger. Group Finance consolidates both sets of numbers into the EXCO dashboard. By the time the dashboard is published, the underlying data is three to five days old. ANG is making active balance sheet and collections decisions on information that is, by design, already out of date when it is first seen.
The reconciliation gap is not theoretical. In Q1 2025, a formal discrepancy incident was logged when the Collections MIS submitted by Group Operations showed a 3% improvement in the 30-plus DPD bucket for the retail mortgage portfolio. The Finance consolidation report submitted the same week showed a 2% deterioration in the same portfolio. Both teams had pulled from legitimate sources. The difference came from cut-off timing, exchange rate treatment, and a data mapping inconsistency between the loan management system and the general ledger. EXCO received two contradictory numbers in the same weekly pack. Resolving the incident took 11 working days and required a joint investigation team. No single person understood both systems well enough to resolve it alone.
The root cause is architecture, not capability. ANG's reporting stack was built incrementally over 10 years, with each new system layered on top of existing ones rather than replacing them. The same underlying data passes through different transformation rules in different systems before appearing as a fact in a report. There are currently 14 manual steps across the three teams involved in producing the Monday pack. Seven of those steps involve copying data between systems. None are logged in an auditable workflow. When a number is wrong, the investigation starts from scratch every time because there is no data lineage trail to follow. ANG does not have a data problem. It has a trust problem, and the data architecture is the cause.
| Report Name | Owning Team | Data Sources | Frequency | Manual Steps | Last Reconciled | Discrepancies This Quarter | EXCO Visibility |
|---|---|---|---|---|---|---|---|
| Collections MIS | Group Operations | LMS, Arrears Platform, Legacy System | Weekly | 6 | 3 weeks ago | 4 | Yes |
| ALM Position Report | Financial Markets | Treasury Mgmt System, Funding Ledger | Weekly | 4 | 2 weeks ago | 2 | Yes |
| EXCO Dashboard | Group Finance | Operations Report, ALM Report, GL | Weekly | 5 | Never formally | 3 | Yes |
| Special Asset Review | Group Operations | Collections MIS, Credit System | Weekly | 3 | 1 month ago | 5 | Partial |
| Liquidity Coverage Ratio | Financial Markets | Treasury System, Regulatory Feed | Daily | 2 | 1 week ago | 1 | Partial |
| Retail Portfolio MIS | Group Finance | Loan Platform, GL | Monthly | 4 | Current | 0 | Yes |
| SME Collections Report | Group Operations | LMS, CRM | Weekly | 5 | 5 weeks ago | 6 | Partial |
| Funding Cost Analysis | Financial Markets | Treasury System, Market Data Feed | Weekly | 3 | 3 weeks ago | 2 | No |
| NPL Movement Report | Group Finance | Credit System, GL | Monthly | 3 | Current | 1 | Yes |
| Islamic Banking MIS | Alliance Islamic CEO Office | Separate Islamic Core System | Weekly | 7 | Never | 8 | Partial |
| Branch Performance Report | Group Operations | CRM, Teller System | Weekly | 4 | 2 months ago | 3 | No |
| Treasury P&L Flash | Financial Markets | Trading System, GL | Daily | 2 | 1 week ago | 0 | Yes |
I want to raise something that has moved from operational inconvenience to genuine business risk. The Q1 discrepancy incident took 11 working days to resolve: Collections and Finance submitted contradictory numbers on the same portfolio in the same weekly pack, and while the investigation ran, the Special Assets Committee met twice with data that had a question mark over it. My team spends an estimated 22 person-hours every week on the manual steps required to produce the Collections MIS alone. That figure excludes the reconciliation loops, the version control issues, and the follow-up queries from Finance when our numbers do not align with theirs.
The fundamental problem is that three teams are producing reports from the same underlying reality but drawing from different systems with different cut-off times, different data definitions, and different transformation logic. There is no single source of truth for even basic metrics like 30-plus DPD or total collections. Every report is internally correct according to the system it came from, and that is exactly the problem. I would like to propose a 60-day data reconciliation and reporting architecture review, with the specific mandate of identifying which system should be the master record for each key metric and what it would take to automate the extraction layer.
Using the report inventory provided, identify which reports carry the highest risk of discrepancy and which teams are most exposed. What are the common root causes across the incidents logged this quarter? Use Copilot to analyse the manual steps, reconciliation status, and discrepancy counts across all 12 reports, then ask it to produce a risk ranking based on three factors: number of discrepancies this quarter, EXCO visibility, and time since last reconciliation.
What would a single-source-of-truth reporting architecture look like for ANG's Operations and Treasury functions? Define the data ownership principles, the automation priorities, and what the EXCO Monday pack should look like when it is no longer built by hand. Use Copilot to help you design the target state: which reports should be consolidated, which data sources should be retired, what the reconciliation governance model should be, and how a discrepancy incident would be detected and resolved in the new model.
Identify the three reports that should be the first to migrate to the new architecture. Define what "one source of truth" looks like as a measurable outcome for EXCO, and what a successful 90-day sprint would deliver. Use Copilot to draft a prioritisation recommendation formatted for the CFO, with a clear before-and-after picture, the business case in numbers, and the decision that needs to be made to unlock progress.
Your working folder contains 2 files for your team:
Opens in OneDrive. Use Incognito / InPrivate mode.
Power Ups are your next step — take what you built today and think about how to bring it back to your real work at Alliance Bank.
Complete your challenge tasks first, then unlock them here.
You have used Copilot to solve a scenario. Now bring it back to your real work at Alliance Bank.
In Tasks 1 and 2, you found that reporting discrepancies are not caused by bad data but by the same data passing through different transformation rules in different systems with no clear ownership decision. You then designed a single-source-of-truth architecture and a reconciliation governance model. Alliance Bank's Operations, Treasury, and Finance teams deal with the same structural problem in every reporting cycle.
Use cases from operations, treasury, and business management teams at banks like yours. React to these: swap, add, or keep.
As a team, pick your top 3, then paste this into Copilot:
Your winning use case needs a great prompt. Paste this into Copilot:
Test it. Refine it. Make it yours.
Draft your adoption pitch: