Where AI returnsthe most leverage
Context & participants
Who was in the room and what we set out to map.
We sat down to map the business start to finish — the history, where you want to take it, and where AI returns real hours rather than adding another tool nobody adopts. Julian opened on how the office came to be and the three-year picture for the first hour, then stepped out; from there the conversation went deep into how the work actually flows, with Paige and Sam.
- Julian Culton — owner. The vision, the control thesis, the build-vs-buy call. (Present for the first hour.)
- Paige King — Operations Manager. How the work runs day to day, where the hours go, and what agents actually feel.
- Sam Schneider — VP Finance. The back-office, commission, and franchise-economics side.
- Jason Cooper — wyth.ai. Facilitating, and running the session through the tool we built for it.
Three people came up repeatedly but weren't in the room: Jeff (managing broker of the Durham office and the deal doctor agents call when a deal hits an obstacle), Gina (operations and paperwork), and Emily Wade (your remote transaction coordinator). Build decisions, you noted, sit with Julian, Jeff, and Gina — independent of any corporate sign-off.
The house you built
What you've made here — identity and edge, in your words.
You started as an independent firm in 2011 and joined Nest for marketing support — after proving to yourself how hard the alternative was. But marketing isn't what keeps your eighty agents. It's leadership, coaching, and a culture you can actually name: nimble enough to change on a dime, structured enough to run like a real company.
“We are a small, independent startup, but we offer the tools of a large company. We've been told we have one of the best tech stacks in the industry — the problem is, they don't all talk to one another.” — Julian, on the identity and the gap this engagement is about
- The edge a competitor can't copy in a weekend: the culture and the collaboration — agents teaching agents, selective recruiting, a managing team people actually lean on.
- The value you under-deliver on today, for lack of hours: frequent one-on-ones and intentional check-ins (two managers for ~65 agents, only a standing annual review), and transaction-coordination support on every deal rather than the ~50–60% Emily covers now.
- Who you serve: your clients are the agents. Make their day faster and their deals easier to win, and the rest follows.
You don't recruit so much as select — the agents who thrive are often career-changers with a strong sphere, taught the Nest way of doing a deal.
(~10–12 agents each)
The lay of the land
Team shape, the stack in daily use, and where work falls between tools. The factual baseline.
Roughly 80 agents across two offices, served by six or seven staff — a deliberately heavy payroll so agents get a high level of service. Jeff is the deal doctor agents call when a deal hits an obstacle; Paige runs operations and onboarding; Gina keeps paperwork on track. Emily Wade coordinates transactions remotely, covering 50–60% of deals. Charlottesville is, in your framing, a publishing house you subscribe to — branding and the eight-a-year Friends of Nest mailers, at roughly 1.45% of gross.
The gap is the seam between all of it. A signed contract gets keyed again and again — the price and address typed into ReChat by hand because nothing reads them off the document, the signed PDF downloaded from email and uploaded into ReChat's unsorted slot, the same data re-entered into LoneWolf and AirTable. And onboarding a new agent runs across many systems — part of the operations work that's about 90% of Paige's day.
Where the hours go
The repetitive, time-eating workflows — with who feels them — and the deal from signature to close.
- Document transmission & re-keying. The agents' number-one complaint is repetition: sign in DocuSign, pull it from email, hand-upload to ReChat, retype the purchase price and address. “Duplicate — just fill it out for me.”
- Contract into the back office. Producing the offer (often through Emily), signature, compliance, then the same data into LoneWolf and AirTable — every hop manual, all the email manual.
- Client & transaction communication. The “where are we” messages run off Emily's proprietary checklist and reminders. Nothing goes out without review — “a layer of control I have not yet relinquished.”
- The Deal Desk. Emily coordinates the deals and Jeff is the deal doctor agents call when one hits an obstacle — it's concentrated in a few people, so deals can stall when any of them is out.
The single-point-of-failure risk surfaced clearly. If Emily is out, deals can stall. If Gina were gone for long, things can break down. The work runs on a few people holding checklists in their heads — exactly the kind of knowledge a Nest-owned layer can capture and back up.
Vision & the skeleton
The three-year picture and the control thesis — sound, and worth building toward.
The vision is a Nest of the Triangle that stays competitive with the Compasses of the world instead of falling behind them — and, in the back of your mind, a business worth more because you own the thing that runs it. The fear is the mirror image: hand your whole stack to one platform and, once 80 agents are locked in, that vendor owns the spine of your business and the pricing leverage that comes with it.
A single all-in-one quote you mentioned would run north of $100,000 a year — a huge share of profit, and a no. The answer you landed on is to own the layer yourself: a proprietary data layer in-house that tools plug into when they're needed, so no vendor gets your data but every tool can still work off it.
“I want to hire someone who develops everything behind the scenes, and we build it out so that eventually we own it locally.” — Julian, on owning the skeleton
This is the right instinct, and it's buildable — on your timeline, alongside ReChat, with no pressure on agents to move until the new layer earns it. The boundary you drew is the one that makes it safe: active-contract data stays on your own box and is never sent to a third-party AI; client and financial details are processed locally. NCREC compliance, ReChat, and DocuSign stay where they are — we build additively, around regulated infrastructure, never through it.
Own the skeleton.
Where the leverage is
Every workflow that surfaced, weighed three ways — the hours it gives back, the edge it builds, and how cleanly it ships without touching regulated infrastructure. The data layer scores highest on value and lowest on discreteness, which is exactly why a clean first brick comes before it.
The eight plays above give your team hours back. These are the client-facing flip side — moves an agent shows a seller or buyer to win the deal. Each needed a Compass-sized budget two years ago; AI can collapse that cost, and on your own captured data they're now buildable in-house.
The trick isn't any one of these. The advantage is owning the data spine and the builds, so Nest can recombine new tricks cheaply and for good, instead of renting each one and re-keying between them.
First-build options
A few strong places to start — each discrete, each returning hours early, each building on what you already run. With a recommended starting point.
The signed-document router
It watches the DocuSign API, reads each executed document, and matches it to the right open deal by property and client — ending the download-from-email-and-re-upload step that tops your agents' complaint list. We ship it in two safe stages: first it surfaces every signed doc alongside the deal it belongs to, read-only, with nothing that can break, so the back office stops hunting through email; then it files them automatically as we confirm the path into ReChat, with anything uncertain held for a quick human check. It rides infrastructure you already pay for, touches nothing regulated, and runs on your own local box. It's the cleanest proof the layer works, and it lays the first brick of the capture-once data layer.
Shape & success criteria
Rough scope, what “working” looks like, and what is explicitly out of scope for the first build.
- Scope: a watcher on the DocuSign API plus a matcher that ties each signed document to an open transaction by property address and client. Stage one surfaces the match read-only — no write, nothing to break; stage two files it into ReChat once that path is confirmed, with low-confidence matches always held for a human. Runs on the local Mac Mini; active-contract data processed on-box, never sent to a third-party AI.
- “Working” means: an agent signs a document and doesn't have to touch it again. It lands in the right file on its own, the manual-transmit step largely drops away for the deals it covers, and accuracy is high enough that the agent's only job is a quick confirm — not a re-do.
- Out of scope: voice capture, contract generation, the full capture-once data layer, the commission engine, and anything that modifies ReChat itself or touches NCREC compliance filing. Those are sequenced, not skipped.
Sequenced path
What comes after the first build, toward the broader vision. The proof earns the rest.
Once the router proves the local-box pattern, voice-to-contract and contact capture is the build you most want — agents speaking a deal or a new contact and having it load into your documents and CRM. Behind it, transaction-coordination assist scales Emily's checklist across every deal and drops the per-deal cost. Those two feed the capture-once data layer — the skeleton itself — into which the LoneWolf and AirTable re-keying finally collapses, with the commission engine riding on top. Each step is additive, on your timeline, alongside ReChat — and if a build proves out, there's a real path to it flowing back to Charlottesville as a solution you led.
The Deal Desk
Opportunity 1 of 4 · its own build & discovery session.
Onboarding & offboarding
Opportunity 2 of 4 · its own build & discovery session.
Training — the “Nest way”
Opportunity 3 of 4 · its own build & discovery session.
Agent dashboard & seller-intent
Opportunity 4 of 4 · its own build & discovery session.
The keys are yours
This plan stands on its own. Before we talk terms, I want your read on it.
Everything here is yours: the map of where the hours go, the ranked opportunities, and the discrete first build to start with. I told your team they'd walk out with an actionable plan, and this is it, usable whether you build it in-house, with another partner, or with me.
What I want first is your take. You know this business better than I ever could, and the plan only earns its place if it matches what you see and want. Tell me where it lands: what's right, what I've weighted wrong, what you'd sequence differently.
Once I have your input and we're aligned on the first build to put in motion, I'll finalize the scope, timeline, and retainer in a separate proposal. No rush on that. The plan is yours to keep either way.