You read the note. Now let's look at your stack.
Twenty minutes. No slides. We'll diagnose where Excel is actually load-bearing in your operation — and whether the problem is solvable with better software or whether you need a data layer underneath it.
Trusted by institutional finance teams across private credit, asset management, and specialty lending.
Jake McFadden
Founder & Managing Partner
Jake McFadden has spent more than a decade at the intersection of structured finance and the technology that runs it. He started inside institutional lending — building pricing models, loan-level data systems, and portfolio analytics at large private credit platforms and across the broader middle-market credit world, where a single broken reconciliation can hold up a deal. He studied Economics and Finance at Princeton, but most of what he knows about this problem he learned sitting inside the operation, watching capable teams lose days to spreadsheets that had quietly become load-bearing.
He builds in the actual stack these firms run on: C# and VSTO Excel add-ins, SQL Server architectures, VBA modernization, Python data pipelines, and Next.js/React applications where the front end matters. His clients are private credit managers, asset managers, and specialty lenders running multi-billion-dollar books on tooling that never caught up to their growth. When you book the call, you talk to Jake — not an associate, not a sales engineer. He has built these systems in production, and he'll tell you straight whether you actually need one.
What we've built.
$20B private credit firm.Four enterprise systems, eight manual handoffs per deal. Built a SQL data layer with custom APIs into Salesforce, Allvue, and ACBS. Reconciliation cycle compressed from 3 days to under an hour.
$30B asset manager.Monthly investor reporting ran through a chain of linked workbooks only two analysts fully understood. Built an automated pipeline from custodian and administrator feeds into a governed reporting engine. The reporting cycle went from three weeks to two days.
Specialty lender, 12 originator feeds.Every new counterparty meant re-keying loan tapes by hand across three systems. Built a normalized ingestion layer with canonical keys and validation at the door. Onboarding a new originator dropped from weeks to an afternoon.
Six signs the problem isn't your software.
Answer honestly. We'll tell you what to do next.
Book the call.
What this call is
- 20 minutes. Camera optional.
- We'll walk through your current stack, identify the seams where data actually breaks, and tell you whether the problem is solvable with better software or whether you need a data layer underneath.
What this call isn't
- A sales pitch. If you don't need what GrayBeck builds, I'll tell you.
- A discovery call that turns into a 90-minute pitch.
- Gated. If a $40K Fivetran license would solve it, I'll tell you that too.