Case Study
ACM Services×Nanonets

From manual batch extraction
to 10× faster invoice processing

ACM Services managed accounts payable for 4,000+ environmental remediation vendors — processing invoice batches 3 times per week entirely by hand. A Nanonets agentic pipeline now ingests batch submissions, extracts all mandatory fields at 98.9% accuracy, assigns GL and cost codes, and bulk-imports directly into Foundation Construction Accounting.

10× Faster Processing98.9% Accuracy4,000+ VendorsFoundation Native
About the customerACM Services

ACM Services is an environmental remediation contractor coordinating work from thousands of subcontractors and suppliers. Their AP team received invoice batches via ZIP archives three times per week — each requiring manual extraction, GL code lookup, and entry into Foundation Construction Accounting before Nanonets automated the pipeline.

Remediation
Specialty
4,000+
Vendors/subcontractors
3×/week
Batch frequency
Foundation
Accounting system

The Problem

ACM Services coordinates environmental remediation projects with thousands of vendors and subcontractors. Invoice batches arrived as ZIP files three times per week — each containing 40–50 invoices from different vendors in different formats. Manual extraction meant hours of work per batch, GL code lookups were done one-by-one, and early-payment discounts were missed because processing was too slow.

  • 40–50 invoices per batch, received 3×/week, each requiring manual field extraction across varied layouts.
  • GL and cost code assignment done one invoice at a time — slowing approval and payment cycles.
  • Early-payment discounts missed due to processing delays — a direct cost to the business.
ZIP Batch
Received 3×/week
Manual
Extract files
Manual
Field entry
Manual
GL lookup
Foundation
Manual import
AP Team
Payment queue
Manual process repeated per batch
47
Invoices per batch (avg)
3×/week
Batch frequency
Stack
Foundation Construction Accounting
Excel

Our AP team used to spend hours on every batch. Nanonets processes the same invoices in minutes and posts them straight to Foundation — it's transformed how we handle subcontractor payments.

AP Manager
ACM Services
ACM Services

The solution

One batch agent. ZIP to Foundation, fully automated.

Nanonets replaced the manual batch loop with one agentic pipeline. Four agents each own one step: ingesting batch submissions and queuing for extraction, extracting all five mandatory fields at 98.9% accuracy across every vendor layout, assigning GL codes and cost codes from the master table, and bulk-importing into Foundation Construction Accounting — with early-payment discount flags surfaced automatically.

Receive Invoice Batch
Extract Fields
Look Up GL Code
Build Foundation CSV
Import to Accounting

Pipeline overview

Vendor Batch
Vendor invoice batch
ZIP / PDF submitted 3×/week
Nanonets
Extract fields
Vendor, invoice #, amount, service
Nanonets
GL code lookup
Vendor Master CSV matched
Nanonets
Transform to Foundation CSV
8-column format validated
Foundation
Foundation Accounting
CSV imported directly
GL Posted
GL posted
Early-payment flag set
Nanonets
Nanonets flag
Unmatched vendor
Manual GL assignment
Vendor master updated
1,040 hrs/yr → near-zero · 98.9% accuracy · Foundation Accounting integrated

Impact

10×
faster processing
Batch invoices automated
98.9%
accuracy
Field extraction rate
4K+
vendors
Covered end-to-end
9/10
satisfaction
Rated by ACM Services
Metric
Before
After
Improvement
Batch processing time
Hours per batch
Minutes per batch
10× faster
Extraction accuracy
Human error-prone
98.9% accurate
AI-validated
GL code assignment
Manual per invoice
Auto-assigned
Rule-based
Foundation import
Manual CSV build
Auto bulk import
Direct integration
Early-payment discounts
Missed — too slow
Flagged automatically
Captured every cycle

AGENT ANATOMY

Nanonets Agent
Autonomous action-taking
After extraction
Agent decides path and acts
Auto-posted
GL matched
Vendor
Batch received
47 invoices detected
ZIP validated ✓
Queue initiated ✓
Nanonets
Fields extracted
VendorEcoClean
Amount$12,450
Nanonets
GL matched
EcoClean → GL 7400 ✓
Cost: REMED-047 ✓
Nanonets
CSV built
Columns8 mapped ✓
FormatFoundation ✓
Foundation
Foundation imported
Rows47
Total$312,450
Foundation
GL posted
StatusPosted ✓
DiscountFlagged ✓
Unmatched vendor
Not in GL master
Nanonets
Vendor not in GL master
⚠ No GL match found
Nanonets
listening…
Nanonets
Flag for manual GL
StatusFlagged
Nanonets
listening…
AP Team
Manual GL assignment
GL code assigned manually
Nanonets
Add to master · loop to CSV
MasterUpdated ✓
Loops to CSV
Amount discrepancy
Total vs line items
Nanonets
Total vs line item mismatch
Sum mismatch ✗
Nanonets
listening…
Nanonets
Flag · AP review
QueueFlagged ✗
Nanonets
listening…
AP Team
Corrected
Amount reconciled
Nanonets
Re-validated · loop to Foundation
StatusValidated ✓
Loops to Foundation

See it run on your process, with your documents.

Start free. No credit card. Or talk to our team about your workflow.