AI agents for manufacturing exceptions

Manufacturing breaks in the gaps between ERP, email, spreadsheets, and people.

OpsCover helps small and mid-sized manufacturers resolve operational exceptions across suppliers, orders, inventory, and production schedules.

The problem

Manufacturers already have ERP systems. But when the normal workflow breaks, the real work still happens manually across emails, spreadsheets, phone calls, and human coordinators.

Supplier delays

A supplier is late, but the impact is buried across purchase orders, sales orders, inventory reports, and customer commitments.

Material shortages

Teams need to understand which jobs, shipments, and customers are affected before the delay becomes expensive.

Manual coordination

Operations teams still read emails, check ERP screens, update spreadsheets, call suppliers, and decide what to do next.

An AI operator for broken workflows

We are not replacing ERP. We are building the AI operator that helps manufacturers recover when the workflow breaks.

01
Ingest

Emails, ERP exports, spreadsheets, purchase orders, sales orders, and production notes.

02
Detect

Identify supplier delays, missing materials, order changes, and production risks.

03
Connect

Match the exception to affected orders, inventory, customers, and production schedules.

04
Act

Draft supplier follow-ups, customer updates, and recommended next actions.

05
Escalate

Send only the highest-risk cases to humans for judgment and approval.

First wedge: supplier-delay exception handling

We start with a narrow but painful workflow: when a supplier is late, manufacturers need to know which customer orders are affected and what to do next.

What the agent does

  • Detects late-supplier issues from emails and order data
  • Identifies affected customer orders
  • Checks related inventory and production data
  • Drafts supplier follow-ups and customer updates
  • Escalates high-risk cases

Who it is for

  • Small and mid-sized manufacturers with 50–500 employees
  • Machine shops and metal fabrication shops
  • Industrial parts suppliers
  • Packaging manufacturers
  • Contract manufacturers

The real workflow is not inside the ERP. It happens in the gaps between ERP data, emails, spreadsheets, supplier updates, and customer pressure.

Founder

Felix Tsai

Founder

Physical-world operations Distributed assets Workflow design

Why this founder

Felix previously worked on large-scale distributed physical operations, where revenue depended on turning messy real-world signals into daily operational decisions.

At Meituan’s shared power bank business, he worked on problems involving device uptime, merchant coordination, location performance, field operations, and business operations across a large physical network.

Manufacturing is a different industry, but the workflow pattern is similar: physical operations break in messy ways, and teams still coordinate across fragmented systems to recover them.

Status

Stage

Customer discovery and early prototype.

Prototype

Lightweight upload-based demo planned for emails, ERP exports, spreadsheets, POs, SOs, and production notes.

Looking for

Small and mid-sized manufacturers willing to share real supplier-delay workflows and test the first prototype.