Inside UniversAIDS: how we trace a bottleneck to its source

Every organisation that asks for AI already knows something is wrong. Reports arrive late, reconciliations eat whole days, customers wait. What it rarely knows is why. The visible problem is usually several steps downstream of the real one, and a system built against the visible problem inherits the real one intact.

UniversAIDS, the UniversAI Discovery System, exists to close that gap. It is the internal platform our consultants use during the Discovery phase of every engagement, and it is built around one idea: a finding is only useful if you can see what caused it and prove where it came from.

What goes in

Discovery starts with the material an organisation already has. We do not ask teams to fill in new templates. UniversAIDS takes in the documents and records that describe how work actually happens.

• Spreadsheets and trackers, including multi-tab reconciliation workbooks
• Standard operating procedures, policies and compliance checklists
• Exports from core systems: ERP, POS, CRM and core banking logs
• Transcripts of structured interviews with leadership and front-line staff

Each source is held in a partition for that engagement. Processing runs only when a senior consultant triggers it, and every run is logged.

From symptoms to a graph

As evidence is processed, UniversAIDS records findings as entities: a bottleneck, a process, a system, a candidate root cause. It then records the relationships between them. A manual reconciliation step depends on a batch import; the batch import fails on a schedule; the failure traces back to a handoff with no sign-off. The result is a graph rather than a list, and the graph is the point.

A list of problems tells you where it hurts. A graph of causes tells you where to cut.

When several symptoms in different departments converge on the same node, that node is a strong candidate for the root cause. Fixing it removes more pain than fixing any single symptom, which is why our implementation recommendations so often look smaller and cheaper than the problem description that started the engagement.

What the consultant does

UniversAIDS proposes; it does not decide. Every candidate root cause is reviewed by a senior consultant, tested against further interviews, and either confirmed, revised or rejected. The graph changes as the evidence does. Only confirmed findings reach the client report, and each one carries its sources.

1. IngestEvidence from documents, systems and interviews is loaded into the engagement's partition.
2. CorrelateUniversAIDS links findings into a graph of bottlenecks, processes, systems and causes.
3. ValidateA senior consultant tests each candidate cause against the evidence and the people who do the work.
4. RecommendConfirmed causes become a costed blueprint for the systems worth building, in the order worth building them.

What the client receives

The output of Discovery is a bottleneck map, an ROI and payback model, a system architecture blueprint and a readiness assessment. Because every finding is traceable, leadership can challenge any line of it and see exactly where it came from. That traceability is what makes the recommendation something a board can approve.