Private preview
Bwat becomes the analyst who can trace every answer back to the evidence
Bwat Analytics is for companies and public institutions whose knowledge is scattered across folders of PDFs, Word documents, spreadsheets, reports, and exports. It is being designed to turn that material into a connected, queryable view of how the organisation actually works.
This is not a generally available self-serve product. Pilot scope, hosting, retention, and model routing are agreed with each organisation.
The system should understand that a supplier in one workbook is the same organisation named in a contract, a payment export, and a project report. Analytics is being designed to extract entities, relationships, events, measures, and provenance into an ontology that can support a living digital twin.
Bring documents, spreadsheets, scans, exports, and selected systems into one governed analytical workspace.
Extract the entities and links that let a question travel across files that were never designed to work together.
Make the source behind a value or relationship inspectable before it is used in analysis or a decision.
Ask an open-ended question, let the assistant gather the relevant evidence, and steer the analysis as the real structure emerges.
Move beyond one-off summaries toward a digital twin that reflects entities, processes, events, and changing measures.
Agree who can access each source and derived result, where they are stored, how long they remain, and which models may process them.
The current Bwat can begin with uploaded material. A pilot adds the connected organisational model and agreed information controls.
“Using the programme documents and spreadsheets I provide, trace each planned output to its budget line, supplier, payment record, and reported delivery. Cite the source location for every link, and list contradictions or missing evidence separately.”
“Review these contracts, operating reports, and payment exports for risks that only become visible when the sources are compared. Build a table with the risk, supporting evidence, affected entity, financial or operational exposure, and the question a human analyst should investigate next.”
Protecting the original files is not enough if extracted text, relationships, search indexes, conversation history, and model outputs are left outside the same controls. The Analytics design treats those derived forms as part of the organisation's information boundary.
The exact technical and contractual controls depend on the deployment. That is why Analytics starts as a scoped enterprise or government pilot rather than a public upload box with broad promises.
Name a recurring question that currently takes analysts days to answer across several sources.
Identify the folders, sheets, systems, ownership, sensitivity, and update rhythm behind that question.
Extract the ontology, connect records, and test known cases with the people who understand the operation.
Add more sources and questions after provenance, access, and answer quality survive real review.
We will map the sources, controls, ontology, and pilot boundary with your analysts, data owners, security team, and decision-makers.
