← All use cases
BNP Paribas Cardif

Automating finance operations

Manual processes in a financial services company, mapped, redesigned and automated, so people could spend their time on work that needs judgement.

120h+of manual work removed every month
↓shorter response times on service levels
3groups aligned: business, IT, vendors

The challenge

Operations teams spent hours every month on repetitive, rule-based tasks. Response times suffered and errors crept in.

The fix had to work across business, IT and vendors, in a regulated financial environment.

The shape of it

automation-hub.internal/overview
Automation hub

Overview

October
Hours saved this month120h
Runs today1,284+12% vs last week
Exceptions3need a person
SLA met98.6%+4.1 pts
Processeshours saved
Invoice matchingAutomated42h
Payment reconciliationAutomated31h
Month-end close checksAutomated24h
Vendor onboardingIn review14h
Expense auditMapped9h
Live runslive
Reconcile batch #48210.8s
Match invoice INV-209311.2s
Close check: accruals2.1s
Match invoice INV-20932sent to review
Vendor pre-check1.6s
Reconcile batch #48220.7s
Payment file export0.4s
Match invoice INV-209331.1s
Reconcile batch #48210.8s
Match invoice INV-209311.2s
Close check: accruals2.1s
Match invoice INV-20932sent to review
Vendor pre-check1.6s
Reconcile batch #48220.7s
Payment file export0.4s
Match invoice INV-209331.1s

Illustrative screens with invented data.

What we did

Discover

Interviews with the people doing the work. Existing data and workflows mapped step by step.

Design

Waste and optimisation opportunities found. Better workflows designed and translated into automation logic.

Build

Robotic process automations shipped. More complex financial changes specified with development teams.

Improve

Requirements for a company-wide automation programme. Today we would add AI agents and n8n to the same toolbox.

What it took

Want results like this for your business?

Tell us what you are building. We reply within one working day with what we would do first.

Tell us your idea
Next use case

A knowledge and rating platform for beer

→