Operations, unblocked.

When a supplier changes a price, misses a delivery, or sends incomplete paperwork, the resolution can spill out of the ERP and into someone’s inbox.

• Context is scattered. Orders, emails, attachments, and company rules need to be pieced together.
• Exceptions become manual work. People check details, ask for clarification, and keep records aligned.
• Local knowledge carries the process. The next step may depend on an experienced operator’s judgment.

Start with one recurring workflow. Bring together the evidence, propose the next permitted step, and verify the outcome.

• Understand the case. Connect the ERP record with the relevant supplier messages, documents, and SOPs.
• Act within clear boundaries. Apply explicit permissions and rules. Ask for approval when a change needs human judgment.
• Show the work. Preserve the evidence, decision, approval, and resulting system change in a traceable record.

The project has positive subject-matter-expert feedback, concept decks and demos, and documented procurement-workflow research.

• A concrete starting point. A BAS procurement case describes document intake, price checks, purchase orders, and delivery follow-up.
• An important boundary. That workflow evidence does not establish demand or economics in North America or Europe.
• The next proof. Validate a recurring exception with a named buyer, historical cases, an agreed baseline, and a pilot decision.

Our thesis: improving AI interpretation and execution capabilities may make some exception-heavy workflows economical to automate.

• Language meets operations. Supplier messages and documents are a candidate entry point for AI-assisted work.
• Control remains essential. Interpretation needs to be paired with permissions, approval rules, and verification.
• The window needs proving. Native agents are a serious alternative. We must find a workflow they cannot serve as effectively or economically.

Our starting hypothesis is a mid-market distributor or light manufacturer with recurring procurement exceptions and a customized ERP environment.

• The operator. Procurement teams dealing with supplier changes, incomplete records, and delivery follow-up.
• The buyer. A procurement or supply-chain leader, with IT and finance involved in integration and controls.
• The expansion path. Earn trust in one workflow before testing adjacent invoice, order, or master-data exceptions.

Customers can use native ERP agents, automation platforms, specialist tools, implementation partners, or their existing team.

• Native ERP agents. We need to prove a useful gap in the chosen ecosystem, rather than assume standard features are insufficient.
• Automation and specialists. We need to show better exception coverage, deployment effort, or total cost for a defined workflow.
• Manual work and custom projects. We need measurable improvements that outweigh integration and ongoing oversight.

Test a paid, fixed-scope pilot followed by an implementation fee and recurring workflow subscription.

• Establish the baseline. Measure case volume, handling time, error rates, and the cost of unresolved exceptions.
• Prove the value. Compare completed cases, operator effort, control quality, and total delivery cost.
• Make delivery repeatable. Reuse connectors, mappings, and policies within one customer segment before expanding.

Can we build a repeatable business around operational work that companies still struggle to automate?

We’re exploring that question from Canada and Germany, with early input from subject-matter experts and a focus on customers in North America and Europe.

Our current priority: customer discovery, a narrow product scope, and evidence from real workflows.

Our financial model will depend on what it costs to win, integrate, and support a repeatable workflow—and what customers will pay for the result.

• Revenue inputs. Paid-pilot conversion, annual contract value, and expansion into additional workflows.
• Cost inputs. Integration effort, AI and infrastructure usage, monitoring, support, and sales cycles.
• The next milestone. Establish pilot economics before presenting revenue forecasts or a fundraising target.