🛠️ BUILD OF THE WEEK

How a 2-person team manages $50M+ and looks like a 10-person shop

A few weeks ago I got off a call with a GP that genuinely mind-blew me. I posted about it on LinkedIn and it took off:

The DMs were the real story. Operator after operator asking the same thing: "How? Walk me through exactly what he does."

So I did. I mapped his entire operation and turned it into a 7-page playbook. It's free, link at the bottom. But first, the part that matters.

He didn't hire more people. He rewired his back office.

Eighteen months ago this GP was doing what almost everyone reading this does. He paid a fund admin. He spent a full week every quarter on a waterfall spreadsheet he couldn't fully explain anymore. He pulled property reports by hand, keyed them into QuickBooks, and wrote investor updates from a blank page.

Then he fired the fund admin and rebuilt the whole thing around AI.

Here's what that actually looks like day to day:

• His property management reports and monthly accounting reports get fed into Claude every week. It finds the discrepancies, flags the questions, and drafts the actual emails to his regionals. "Why is leasing soft here?" written for him, ready to send.

• It scrapes live rent comps so he sets rents off actuals, not guesses.

• When he sits down to write quarterly updates, the system already knows the last three months of context. It drafts 98% of the update in five minutes.

• His pitch decks, investor guides, LP return projections, and underwriting models all run through the same setup. One of those models passed an institutional investor's sniff test.

By the numbers

What rewiring the back office around AI actually bought him:

• Monthly investor updates: 98% drafted in 10 minutes

• Waterfall + distributions: a full week in Excel → effectively one click

• Time on back office: 60% → 10%

• The fund admin bill: gone

• The team: still two people

That last 90% of his time? It goes to sourcing deals and raising capital. And it shows. He's closing deals and raising money at a pace most groups in his market can't touch, to the point where his next hire is a second capital raiser, not an operations person. The back office stopped being the bottleneck.

That's the whole unlock. Using AI as a chatbot saves you minutes. Rewiring your systems around it replaces entire functions.

What the old way actually costs you

If you're managing $50–300M across 10–30 SPEs, here's what the spreadsheet really costs every month:

• $1,500–3,000/mo for a bookkeeper or part-time controller who reconciles GLs and fixes the spreadsheet when it breaks

• $800–1,500/mo for a CPA firm that re-keys what the bookkeeper already keyed

• 10–15 hrs/mo writing investor updates by hand

• 3–5 hrs/period on a waterfall in Excel nobody fully understands anymore

• 2–4 hrs/mo answering "can you resend my K-1?" for the tenth time

Call it $4,000–8,000/mo plus 20+ of your own hours. And that's when everything goes right.

One wrong cell in a waterfall and you've sent incorrect distributions to 200 investors. One late K-1 and your inbox lights up. LPs notice. They talk. And when your next deal comes around, they're suddenly "overcommitted."

Your back office isn't a cost center. It's your reputation. Your next raise depends on it.

The new way, in six steps

This is the workflow inside the playbook. The whole thing runs in ten minutes instead of four hours.

1. Property data flows in automatically. Your PMS connects to your fund admin platform. No manual pulls, no CSV uploads.

2. Distributions calculate in minutes. Waterfall logic is configured once. The system knows the distributable amount, runs the waterfall, shows the breakdown by investor. You review and approve.

3. Waterfalls run without spreadsheets. Documented, auditable, consistent. No fat-finger errors.

4. Investor notices generate themselves. Personalized per investor, ready to send the moment you approve.

5. Quarterly updates write themselves. You've been feeding data in all quarter. AI drafts 90–95% of the narrative. You edit and send.

6. Journal entries export to QuickBooks. Clean data for your CPA. No re-keying.

One source of truth. Systems that actually talk to each other.

📋 WHAT'S IN THE PLAYBOOK

I broke the entire system down into 7 pages. Here's what you get:

• Why your LPs already know your back office is messy (and what that's quietly costing you on your next raise)

• The exact 6-step workflow that cuts distribution time from 4 hours to 10 minutes

• The full tech stack the $50M GP uses, six pieces, connected not stitched

• How to look like a 10-person team with 2 people

• A self-assessment checklist to score how AI-native your back office actually is right now

The one rule that ties it all together: AI doesn't fix fragmented data. It amplifies whatever's already there. Connect the systems first. Then put AI on top.

That's the difference between the GPs closing three deals a quarter and the ones stuck on one.

👉 Download The AI-Native GP Playbook

Want me to map it to your stack?

The playbook shows you what to build. If you'd rather see it running on your own deals, that's what I do.

Book a 30-minute call and I'll map your current stack (Rent Manager, Yardi, AppFolio, PropertyWare, QuickBooks, whatever you're on) and show you exactly what's automatable. Most GPs are live within 4–12 weeks depending on their setup.

Get a free audit

The gap between AI-native operators and everyone else is widening fast.

Pick the side you want to be on.