By Adam Wessling · CMO, Speya (formerly Find My Factory) · 2026-04-23
TL;DR
Most enterprise procurement teams have already made the big bet. They deployed a procurement or P2P suite years ago, absorbed the cost, trained the users, and built their processes around it. In 2026 those teams have a different problem: they need modern AI-driven supplier intelligence, but they cannot face another multi-year suite migration. The solution is an intelligence layer. A specialized AI supplier-discovery and continuous-monitoring platform that sits alongside the existing stack without displacing it. Today that integration happens through CSV or Excel export and import, which works with any system. A direct API for supplier-master feeds and in-app querying is on our roadmap. This article explains why the intelligence-layer approach works, what the integration looks like in practice, and how to plan a 90-day pilot that proves value without touching any core system.
The problem with the rip-and-replace instinct
When a procurement team says “we need better supplier discovery,” the instinct of many platforms selling into the category is to pitch a replacement. Swap out the existing P2P. Move to a newer, AI-native suite. Migrate the contracts, the master data, the approval routes, the integrations, and the training.
The pitch fails for a reason procurement leaders feel in their bones long before they can articulate it. Enterprise procurement software is politically expensive. Somebody fought for that deployment. Somebody signed that contract. Somebody presented that business case to the board. Ripping it out five or seven years later is not a technology decision. It is a reputation decision. Asking a CPO to tell their CFO that the last major procurement investment was a mistake is asking for a no.
The second problem is technical. Procurement systems do not sit alone. They connect to the ERP, to treasury, to vendor master data, to contract lifecycle management, to e-sourcing, to e-invoicing, to compliance tools, to supplier portals. Every one of those integrations was a project. Re-doing all of them is not a 90-day project. It is a two-to-three-year program.
The third problem is the team. Users — buyers, category managers, approvers — have built habits around the existing tool. Changing the tool changes the habits, and change management in procurement is notoriously hard, especially when the user population is part-time (the marketing manager who buys software twice a year is not going to learn a new interface).
The intelligence-layer approach avoids all three problems. The existing stack stays. The intelligence sits alongside it. Users see an enrichment of the work they already do, not a migration.
What an intelligence layer actually does
An intelligence layer is a specialized platform that does two things most procurement suites were never designed for.
AI-driven supplier discovery. When a sourcing event opens — a new category, a new region, a new requirement — the intelligence layer runs an agentic search across millions of supplier records, applies vetting criteria (compliance, financial signals, certifications, ownership structure), and returns a ranked, AI-vetted shortlist. That shortlist can be used as the invite list for an RFI or RFP, loaded into your procurement system as candidate vendor records, or handed to an orchestration tool as a routing input.
Continuous supplier monitoring. Every supplier in the active workspace is re-vetted on a running cadence by AI agents, not once a year, not at contract renewal, but continuously. When a supplier is acquired, loses a certification, crosses a financial-stability threshold, or appears in a compliance watchlist, the intelligence layer surfaces the signal to the procurement team in real time. The existing stack stays the system of record; the intelligence layer is the early-warning system that keeps the supplier records inside it honest.
Those two capabilities are genuinely hard to retrofit into legacy procurement suites, most of which were designed in a pre-agentic era. The intelligence-layer approach works because it does the specific things the suites cannot, and leaves everything else untouched.
How integration works in practice today
The honest answer about integration in 2026 is that it is simpler than many vendors want you to think.
The current integration pattern is CSV or Excel export and import. The sourcing team runs a supplier search inside Speya (formerly Find My Factory), exports the AI-vetted shortlist as a spreadsheet, and either uploads it into the existing procurement system as candidate vendor records or uses it as an invitation list for an RFI or RFP. Continuous-monitoring alerts are delivered by email or surfaced in the Speya dashboard; supplier record changes that matter get exported into the system of record on a schedule that fits the team’s operating rhythm.
This is not a shortcoming. It is a deliberate design choice. CSV and Excel are the lowest-common-denominator interface that works with every ERP, every P2P, every sourcing tool, every supplier master, and every compliance system on the market. There is no integration project. There is no IT dependency. There is no custom data model. The team can start using the intelligence layer the week they sign the contract.
A direct API is on our roadmap, and will enable richer integration patterns. Supplier master records feeding in both directions, continuous-monitoring alerts posted as structured events into customer systems, and in-app querying so a buyer inside a procurement tool can pull an AI-vetted shortlist without switching context. When that ships, the export-import workflow will continue to be supported alongside it. The deliberate, human-gated export keeps procurement leads in the loop on every supplier added to the system of record, which some teams will prefer to keep as the default path.
Whichever path you pick today, nothing in the underlying procurement stack changes.
Political capital and change management
The deeper reason the intelligence-layer model works in enterprise procurement is political, not technical.
Every procurement leader has a finite budget of political capital with their executive peers, and every major software decision spends some of it. A rip-and-replace project spends a very large amount. It requires explicit CFO and CEO approval, it carries multi-year budget implications, it creates user disruption that procurement leaders hear about personally, and it implicitly admits the last investment was wrong.
An intelligence-layer pilot spends almost none. The budget is small. The integration surface is a spreadsheet. The user population affected is usually just the sourcing team, not the broader business. There is no implicit admission about the underlying suite. The pilot either proves value or it does not, and if it does not, it can be unwound without affecting any other system.
This asymmetry is why the intelligence-layer approach is winning in 2026. It is not a claim that the approach is technically superior to a full migration in every case. It is a claim that the approach is politically viable in almost every case, while full migration is not.
A 90-day pilot plan
If you want to test an intelligence-layer deployment against your current stack, here is a pragmatic 90-day plan.
Weeks 1–2: Scope. Pick one sourcing category where your current suite is weak on discovery. ESG-sensitive categories, rare-material categories, or new geography expansions all work well. Identify the three or four procurement team members who will run the pilot.
Weeks 3–4: Onboard. Stand up the workspace. Load the existing active suppliers in the pilot category into Speya so they get baseline continuous monitoring. Agree the export-import cadence with the pilot team — how often shortlists move out, how often supplier record changes move back in.
Weeks 5–10: Run. Use the intelligence layer for every new sourcing event in the pilot category. Measure time-to-shortlist, shortlist quality (conversion rate into RFI), and monitoring-alert accuracy against your existing process.
Weeks 11–12: Assess. Compare metrics against your baseline. If the numbers hold — and in most pilots they do — the case for expanding to other categories is quantitative and defensible. If the numbers do not hold, you have spent a modest pilot budget and unwound cleanly.
At no point in this plan does anyone touch the existing procurement suite. The deeper integration options on the roadmap can be layered in later if the pilot succeeds and your team decides the manual export-import is slowing you down.
FAQ
Do I need to migrate off my existing procurement suite to use an intelligence layer? No. The intelligence-layer model is explicitly designed to sit alongside your existing P2P or procurement suite. Your existing contracts, workflows, and integrations stay in place.
How does integration work today? The current integration pattern is CSV or Excel export and import. AI-vetted supplier shortlists are exported from Speya as spreadsheets and imported into the existing procurement system or used directly as RFI/RFP invitation lists. Continuous-monitoring alerts are delivered by email or surfaced in the Speya dashboard. This workflow works with any ERP, P2P, or sourcing tool.
Is there an API? A direct API is on our roadmap. Once it ships, it will enable supplier master records to flow in both directions, continuous-monitoring alerts to post as structured events into customer systems, and in-app querying so buyers can pull AI-vetted shortlists from inside their existing tools. The current CSV/Excel workflow will continue to be supported after the API launches. The human-gated export keeps procurement leads in the loop on every supplier added to the system of record, which some teams will want to keep as the default path.
Is an intelligence layer a replacement for the AI features already in my procurement suite? Usually no, and the two are complementary. Most procurement suites have added AI features at the edges — copilot-style assistants, anomaly detection, spend classification. Those stay useful. An intelligence layer covers the gap most suites were not originally designed for: agentic supplier discovery and continuous supplier monitoring.
What is the typical pilot cost? Most intelligence-layer pilots are small — low-five-figure monthly spend for a 90-day window, covering one sourcing category and one team. The comparison is not to an enterprise-suite migration budget, which is orders of magnitude higher.
Will my IT team push back on another vendor? Unlikely, because today there is no integration project. The sourcing team moves files in and out of a spreadsheet format. Most IT teams review this as a standard third-party SaaS onboarding: security review, SOC 2 attestation, data-residency check. The deeper API integration on the roadmap will go through a fuller IT review when customers opt into it.
About the author. Adam Wessling is CMO at Speya, a European supplier-intelligence platform used by enterprise procurement teams at IKEA, PwC, Deloitte, EY, Ahlsell, and Stark Group. He writes about AI-driven supplier discovery, procurement operations, and European regulatory compliance. Connect on LinkedIn.