Finance ops · Production

Jump replaces QuickBooks Online with Rillet's AI-native ERP and closes the books in under one week as a team of one

The problem

Jump ran on QuickBooks Online with external bookkeepers producing a 20–30-day close cycle, no native Stripe or Ramp integration, poor customer-level visibility, and a single controller expected to handle the workload of four people with no time for strategic work.

First attempt

A traditional ERP approach (exemplified by a prior NetSuite implementation) took over six months to stand up, required modules and add-ons that reduced automated workflows, and introduced sync issues and maintenance overhead with each external tool — leading Jump's controller to reject that path entirely.

Workflow diagram · grounded in source
1
Stripe payments auto-reconciled
integration
“Rillet's Stripe integration auto-reconciles 99% of thousands of monthly transactions without Brock touching them”
2
Ramp T&E synced in real time
integration
“The Ramp integration syncs in real time and maps cleanly to the GL”
3
AI agents auto-code transactions
ai_action
“With Rillet's agents and matching principles, only about 5% of his monthly transactions require manual coding”
4
Manual review of exceptions
human_review
“95% of monthly transactions auto-coded, only ~5% require manual review”
5
Continuous close delivered
output
“Close shortened from 20-30 days to under one week, with revenue closed on day 1”
Reported outcome

After implementing Rillet, Jump's close shortened from 20–30 days to under one week with revenue closed on day 1, one controller now handles the equivalent of four headcount, 95% of transactions are auto-coded, and the Stripe integration auto-reconciles 99% of thousands of monthly transactions.

Reported metrics
Close duration20-30 days to under one week
Revenue close dayrevenue closed on day 1
previous close duration (QBO)20-30 days
Capacity multiplier4x capacity. 1 person easily doing the job that would require 4
Show all 9 reported metrics
close duration20-30 days to under one week
revenue close dayrevenue closed on day 1
previous close duration (QBO)20-30 days
capacity multiplier4x capacity. 1 person easily doing the job that would require 4
transactions auto-coded95%
transactions requiring manual review~5%
Stripe transactions auto-reconciled99%
manual expense coding time replacedhours of manual expense coding
prior ERP implementation timeover six months
Reported stack
RilletQuickBooks OnlineStripeRamp
Source
https://www.rillet.com/case-studies/jump-replaced-qbo-they-now-close-the-books-in-under-two-weeks-as-a-team-of-one
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

After implementing Rillet, Jump's close shortened from 20–30 days to under one week with revenue closed on day 1, one controller now handles the equivalent of four headcount, 95% of transactions are auto-coded, and th…

What tools did this team use?

Rillet, QuickBooks Online, Stripe, Ramp.

What results were reported?

Close duration: 20-30 days to under one week; Revenue close day: revenue closed on day 1; previous close duration (QBO): 20-30 days; Capacity multiplier: 4x capacity. 1 person easily doing the job that would require 4 (source-reported, not independently verified).

What failed first in this deployment?

A traditional ERP approach (exemplified by a prior NetSuite implementation) took over six months to stand up, required modules and add-ons that reduced automated workflows, and introduced sync issues and maintenance o…

How is this finance ops AI workflow structured?

Stripe payments auto-reconciled → Ramp T&E synced in real time → AI agents auto-code transactions → Manual review of exceptions → Continuous close delivered.