What is — right now.
One live picture of your business, pulled from the graph.
Every number traces back to the source that wrote it, the minute it landed.
Three to five hours of weekly reconciliation — gone.
You are spending three to five hours a week reconciling spreadsheets, cross-checking dashboards, and chasing down which version of which number is current. That is not an ops problem. That is lost revenue, and it compounds every week you tolerate it.
Worse: when a customer exercises their GDPR Article 15 right of access and asks you to show them everything you hold on them, can you actually do it? Right now? In seconds? If the answer is "we'd need to check a few systems," you have a compliance exposure that gets more expensive every month the ICO tightens enforcement.
wot.now fixes both problems at once. Your operational data -- sales, inventory, leads, finance, headcount, whatever matters -- lives once, in a content-addressed graph. Every update from any source (CRM, accounting, manual entry) becomes a signed thought. The dashboard reflects it the moment it lands. One screen. One truth. Every number auditable to source.
Every environment is different, and we know that. But data and its relationships are the one constant, and that is what we do. Whatever your data looks like today -- however many spreadsheets, however tangled -- we have seen worse. We schema-discover it, connect it, and give you one screen that tells you where you are right now.
If you're a founder or ops lead with operational data scattered across CRM, accounting, scheduling, and a stock count that lives in the founder's head — this is built for you. BuyMaterials is the live example: UK construction-materials marketplace, DTS norn running on Azure today, real production traffic, real receipts. Our pro-bono work with British Comedy Guide covers 7.86M rows across 363 tables — a cultural-heritage archive, fully schema-discovered and connected. If that shape of pain is yours, you're in the right place.
BuyMaterials: bypassing Microsoft's JSON-to-SQL impedance on the way to Power BI.
BuyMaterials — UK construction-materials marketplace — went all in on the Microsoft analytics stack: Cosmos DB for the operational JSON documents, Synapse as the warehouse, Power BI on top, detailed semantic models built out over months.
The problem: Cosmos stores JSON natively and handles it well. SQL Server — Microsoft's own relational engine, which Microsoft's recommended pipeline forces you to route Cosmos through to reach Synapse and Power BI — does not handle JSON properly in its own SQL language. You have to sidestep Microsoft's failings in their own product.
The failure mode is brutal: a customer types the wrong kind of quote into a Cosmos document and the entire downstream pipeline breaks. The semantic models inherit the damage. The dashboards show numbers that already diverged from the source — or don't show at all, because the pipeline failed somewhere between Cosmos and SQL and nobody noticed until Monday morning.
With wot.now: the BuyMaterials DTS norn runs on Azure, reads Cosmos directly, and lands each document as a signed thought. No JSON-to-SQL round trip. No Synapse impedance mismatch. No semantic-model drift. Power BI keeps doing what Power BI does — but every figure now has a because-line back to the Cosmos document that produced it.
A customer submits a GDPR Article 15 subject access request. The ops team queries the graph once — not Cosmos, then the SQL mirror, then Synapse, then Power BI's model cache — and returns a complete, signed export of everything held on that customer in seconds.
The broken step bypassed. Receipts on every figure. The Microsoft stack left in place where it still works.