The Knowledge Graph That Remembers Every Supplier Relationship Across 10 Years of Invoices
Persistent Entity Relationship Mapping in Financial Document Intelligence
---
An auditor asks which suppliers account for more than 15% of your total spend. That question has a definitive answer — buried across 10 years of invoices in your accounting system. Your finance team says they'll have it in two weeks.
The data was always there. The time is the cost of having it formatted for filing, not for intelligence.
---
The Compliance Obligation Nobody Reads Twice
French law requires businesses to retain invoices for 10 years under Article L. 102 B of the Book of Tax Procedures — the legal obligation is clear, the archive is maintained, the filing is organised. Those same businesses then spend two weeks manually reconstructing supplier histories when auditors arrive, because the 10 years of retained documents were archived for compliance, not structured for questions.
Storing a PDF for compliance is not the same as extracting what it contains. The archive costs money in server space. Extracting everything it contains is what turns that storage into intelligence.
Your accounting software knows what you paid and when. Stralevo's Knowledge Graph knows who you paid it to, under what conditions, and across every transaction you've ever processed.
---
What Gets Built as Invoices Arrive
Each invoice processed adds three types of intelligence to the Knowledge Graph. Entity nodes — the supplier's legal identity, their contacts, their product codes, their delivery addresses. Relationship edges — how this supplier connects to the purchase orders, delivery notes, and bank payments across your history. Temporal data — how the relationship changed year on year: price, volume, payment terms, contact changes.
SmartHub, the connection-mapping layer inside Stralevo, builds these links automatically across document types. An invoice connects to the purchase order that generated it, the delivery note that confirmed receipt, and the bank payment that settled it — all joined in a persistent map that grows with every document added. When you ask "what is our complete relationship with Supplier X?", SmartHub searches through that map in seconds rather than requiring a human analyst to rebuild it from scratch.
Entity duplication — ACME Corp, Acme Corporation, ACME SARL appearing as three different vendors in three different invoices — is resolved automatically. No manual deduplication. No reconciliation spreadsheet. The same vendor appears as one entity across the full document history.
---
What a Decade of Invoice Data Actually Contains
A manufacturing company with 400 active suppliers discovered that 60% of its critical component spend flows through three vendors — a concentration risk invisible in individual invoice views. That discovery required no new data collection. It required making existing data queryable.
An accounting firm partner running a year-end audit located every warranty obligation across seven years of client invoices in 11 seconds. The data had been filed correctly for seven years. Nobody had ever made it searchable. One Knowledge Graph query eliminated two days of manual document review.
Across the 10-year invoice archive of a mid-market French business: every price-per-unit charged by every supplier, across every product, in every quarter. Every payment term negotiated or adjusted over that period. Every supplier contact change — which often precedes changes in pricing behaviour or service quality. Every volume commitment and whether it was honoured. The answer to every supplier question the board has ever asked is already in there.
Nothing has to be newly collected. Everything has to be newly queryable.
---
The Information Asymmetry in Supplier Negotiations
Before every significant negotiation, your supplier's account team reviews your complete purchase history with them. They know your volume trends, your payment patterns, your price sensitivity. They arrive informed. Your procurement team typically arrives with a recent export from the accounting system — date, amount, description — and no historical context.
Flipping that asymmetry is what the Knowledge Graph does in practice. A CFO who can answer "we've increased our volume with this vendor by 40% over three years, and their prices have risen 8% annually" is in a different conversation than one who cannot. That 8% annual increase on a 200,000-euro supplier relationship means you pay 272,000 euros in year five instead of 200,000 at the rate you started — a 72,000-euro annual gap compounding silently, invisible in flat invoice exports until someone runs the cross-year analysis.
Overcharges discovered at month 36 do not refund themselves. The Knowledge Graph doesn't reverse past losses — it closes the window on future ones.
---
Three Risk Vectors That Accounting Software Misses
Supplier concentration risk arrives not as a boardroom decision but as 200 individually sensible procurement choices — each with a good reason, none of which was assessed for its cumulative effect on total spend distribution. By the time 40% of critical spend flows through one vendor, the risk is already structural. The Knowledge Graph surfaces concentration before it becomes dependency.
Price drift operates the same way. A 2% annual increase looks like rounding error in any single invoice period. Across eight years of data, at 200,000 euros of annual spend, that same contract costs 34,000 euros more per year in year eight than it did at the original rate — a gap that exists in your archive right now, sitting across hundreds of filed invoices, invisible until someone maps the price history.
Relationship changes signal risk before financial changes appear. A new contact at a supplier often precedes renegotiated terms within six months — the Knowledge Graph flags the change the moment it appears in a new invoice header, so the finance team knows before the renewal conversation starts rather than after.
---
What Systematic Supplier Intelligence Changes
CFOs typically cannot answer which five suppliers represent more than 50% of total spend from memory or from their accounting system in under three minutes. That knowledge exists in archived invoices. Reaching it currently requires a manual project that most teams run once a year, if at all — not because CFOs don't want the information, but because extracting it from flat archives takes longer than the timeline allows.
When supplier intelligence is queryable in seconds rather than days, the first-order effect is faster responses to auditor and board questions. The second-order effect is that those questions get asked more often — by the CFO, not just the auditor. A compliance officer who needs to pull every transaction tied to a supplier under regulatory scrutiny — 847 invoices, 23 contracts, complete history — currently treats that as a multi-week project. With the Knowledge Graph already built, it is a single query returning a sourced answer in minutes. Procurement decisions that previously waited for a formal quarterly review become data-driven at the point of decision.
M&A due diligence firms charge 50,000 euros or more to manually reconstruct supplier relationship profiles for acquisition targets. That premium exists entirely because the data is already there — in years of invoices — and not structured to be read. Organisations that can answer supplier questions themselves, in minutes rather than weeks, are not dependent on that premium service for their own operational intelligence.
---
Twelve Months From Now
Customer relationship management software became standard infrastructure once organisations realised they were losing revenue by not tracking customer history. Supplier relationship intelligence is following the same curve. The accounting firms already deploying Knowledge Graph-powered supplier analysis for client reviews are setting a standard their peers will scramble to match — and the standard spreads from the firms that deploy it first to every client they bring onto the platform.
Upload the last three years of supplier invoices. Run one Knowledge Graph query against concentration and price trends. What comes back is not a report someone built for you — it is your own invoice data, finally structured to answer the questions it always contained.
Keeping invoices for 10 years is a legal obligation in France. Using them to understand your supplier relationships is an advantage. Those are not the same thing — and for most finance teams, only one of them is currently happening.