Private preview
Built from the files you already have.
Contracts in one folder, payments in a spreadsheet, delivery reports somewhere else. Bwat Analytics reads all of it, works out when a supplier, a project or a payment in one file is the same one in another, and builds an ontology you can question in plain English. Every answer points back to the row, page or paragraph it came from.
Analytics is in private preview. Hosting, how long data is kept, and which AI models may read it are agreed with each organisation.
Did we pay for any work the delivery reports never mention?
An ontology is a map of the things your organisation deals with, like people, suppliers, projects, payments and places, and how they connect. Analytics builds it from your documents, notices when a name in a workbook is the same supplier as one in a contract, and keeps a link to the source for every connection. As new reports and exports arrive it updates, which is what turns it into a living digital twin.
Which suppliers were paid in March for work no delivery report mentions?
Two. Supplier A was paid UGX 84M on 12 March for grading work that no report covers 1 2. Supplier D was paid twice for the same invoice 3.
Documents, spreadsheets, scans, exports and chosen internal systems, behind one set of access rules.
The people, suppliers, programmes, payments and places in your records, and how they link up.
Open the source behind any figure or connection before you use it in a decision.
Ask something open-ended and let Bwat gather the evidence, then steer as the picture forms.
Go past one-off summaries to a model of how the organisation runs, updated as things change.
Agree who can reach each source and result, where it's stored, how long it's kept, and which models may read it.
You can try these today with files you upload. A pilot adds the connected model and the access 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 and missing evidence separately.”
“Review these contracts, operating reports and payment exports for risks that only show up when the sources are compared. Build a table with the risk, the evidence, the entity affected, the likely financial or operational exposure, and the question an analyst should look into next.”
Locking down the original files isn't enough if the extracted text, search indexes, chat history and answers sit outside the same controls. Analytics treats all of those as part of your data.
Exactly how that works depends on where it's hosted and what's in the contract, which is why Analytics starts as a scoped pilot and never as a public upload box.
One your analysts ask often, and that currently takes them days across several sources.
Which folders, sheets and systems hold the answer, who owns them, and how sensitive they are.
We build the ontology and test it on cases your team already knows the answer to.
Add sources and questions once the answers, the sourcing and the access rules hold up.
We'll work through the sources, the controls and the pilot's scope with your analysts, data owners and security team.
