Thirteen changes to a live lending system. Zero incidents.
Several orgs, a loan system, a nightly transfer that fails every other week, and a backlog that waits months because every release is scary. We ship into live lending systems in small governed steps: a test environment, a sign-off, a release window you choose, a rollback tested beforehand. Named case: GoodLeap.
GoodLeap · Financial Services Cloud, Sales Cloud, Service Cloud, MuleSoft, ICE Encompass, Figure
"The nightly transfer broke again."
Twenty thousand records between the loan system and the orgs, failing silently every other week. Nobody finds out until the loan officers do, in the morning, and the fix is a person with a spreadsheet.
What they ask us on the first call
- How do we find out a transfer failed before the loan officers do?
- Who can roll back last night's release, and how fast?
Work that used to take 6–12 months now ships in weeks.
Read the caseAt GoodLeap the transfers moved to an integration backbone across Salesforce, MuleSoft, Figure and Encompass, released in governed steps. Thirteen releases since, zero critical incidents, one rollback in the period.
"Loan officers assemble the file by hand before every call."
The application is in the LOS, the status in another system, the last contact in an inbox, and marketing attribution gets corrected by hand at month end. The officer spends the first ten minutes of every call finding out what already happened.
What they ask us on the first call
- Can an officer see the application, the LOS status and the last contact on one screen?
- Why does marketing attribution still get fixed by hand at month end?
Work that used to take 6–12 months now ships in weeks.
Read the caseThe first GoodLeap result was a two-week win: attribution set automatically at Opportunity creation. Then dynamic scripting, claim queues and role-based priority views for the front line.
"A two-day change takes a quarter to reach production."
Not because the work is hard. Because the last release broke something on a Friday and now every change waits for a window nobody wants to own. The backlog grows, and the internal estimate for anything is six to twelve months.
What they ask us on the first call
- Why does a two-day change take a quarter to reach production?
- Can we ship every two weeks without breaking Friday?
How a release into a live lending system runs
- Built and tested in a full copy of the system, then by your process owner
- Change record: requested by, built by, tested in, approved by, rollback
- A release window you choose, a smoke test, a week of watch
- Go/No-Go before every release; a rollback tested beforehand
At GoodLeap: four projects in the first four weeks, eight in nine, from an internal baseline of six to twelve months per initiative.
Three ways in. Each priced before the call.
The Health Check first if nobody can say what the system costs. A named win if everyone already knows what is broken.
Health Check, five days, $1,500
One production org. Several orgs or connected instances are a Current State Review, $4,500–$9,000, and we say which on the call.
Nightly transfers that stop breaking
Done means: the 02:00 transfer has run clean for ten working days and someone is paged before the officers notice when it does not. Built first at GoodLeap.
Managed Delivery, a result every two weeks
Small governed streams, the same named team, a release every two weeks with a rollback path. How GoodLeap went from months to weeks.
Systems we have connected for lenders: Figure, ICE Encompass, Insellerate, Verse, Mogli SMS, Genesys Cloud CX, Experience.com, Domo, Databricks, MuleSoft Anypoint and Composer.
What clients in lending and mortgage say.
No public review from a lender yet: GoodLeap is a current client. The case is documented above, and the release records are available under NDA.
Before the first call.
Nightly transfers between Salesforce orgs: what we learned at GoodLeap
A 20,000-record nightly transfer that failed every two weeks, several Salesforce orgs, MuleSoft, Platform Events and a release process. Thirteen governed releases to a live lending system, one rollback, no incidents.
Read After go-live · 4 minWhy Salesforce projects fail two weeks after go-live
Most Salesforce projects do not fail at go-live. They fail when the senior team rotates out, changes go to production without a release process, and the vendor is the only one who understands the org. Three causes, two real orgs, one checklist.
Read Health Check · 4 minWhat a Salesforce Health Check report contains, page by page
The structure of a Salesforce Health Check report: the Org Health Score, six scored areas, priced findings, AI readiness and the 30/60/90-day roadmap, and what the five days look like for your team.
ReadQuestions from lending and mortgage firms
Do you work inside our loan origination system?
We have several orgs. Where do you start?
How do you release into a live lending system without an outage?
Can you show a money result for lending?
Where is the team?
Make your Salesforce earn.
Five days. Every problem priced, a plan for the next 90 days. $1,500, credited in full against the fix.
We reply within one business day with a kickoff slot. Prefer a 30-minute call first? Request two slots.