Things people ask first.

Pricing depends on which engagement option you pick — Full Custom (you own outright, in your market), Design Partner (half the upfront fee, I retain the right to commercialize a generalized version later), or Build & Host (reduced upfront, monthly recurring). All three are discussed on the call, with a one-page fixed-scope, fixed-fee proposal in writing within a few days. See the engagement options →

Design Partner is approximately half the upfront fee of a full custom build. In exchange, I retain the right to commercialize a generalized version of the pattern for other operators — without your data, customer list, or proprietary configuration. You get a perpetual license to the version built for you, including configuration unique to your portfolio, plus a written carve-out preventing me from selling the same implementation in your specific market.

If you have at least one workflow eating more than two hours a week and you can describe it clearly, you're not too small. The sweet spot is 4–20 properties. Useful systems can be built for operators with one well-loved property, but the ROI math gets thin below four. The honest test: if a single workflow — guest messaging, pricing, reviews, maintenance routing — is eating 8+ hours of your week or pushing you toward hiring a VA, a custom build pays for itself inside the first quarter on labor alone, before counting recovered guest experience or earlier issue detection.

No. Knowing your workflow is the prerequisite — not knowing the technology. If you can describe what currently happens and what should happen, I can build the system. I'll teach you what you need to run it confidently as part of the engagement, including how to recognize when the agent is uncertain versus when it's confident, how to add or correct house-rules content over time, and how to read the weekly rollup the system produces. The technical layer is mine to keep working. Yours is the operating layer — and you already have it.

Every system I build has a human review layer for anything material — guest comms, payments, money decisions. The agent drafts. You approve. As it proves itself on specific task types, we expand its autonomy in measured steps. Nothing ships that can quietly fail on you. When the agent is uncertain, it flags rather than guesses — the message lands in a review queue with the relevant context attached, so you can decide in seconds instead of hunting through threads. The system also logs every action it takes, so an after-the-fact audit is one click, not a forensic exercise.

Almost never. Most of what I build sits on top of your existing tools — Hospitable, Hostfully, AppFolio, QuickBooks, Gmail, your spreadsheets — and connects them in ways the tools themselves don't. We design around your stack on the call. You don't migrate, you don't switch PMS, you don't retrain your cleaner. If the discovery call reveals that the right answer is "your current PMS already does this, you just haven't turned it on," that's the recommendation — and you save the build fee.

By default, in your accounts on your infrastructure, under your control — your Hospitable, your Airbnb, your QuickBooks, your Gmail. If you'd rather I host the system (the Build & Host engagement), your data lives on infrastructure I run under a written agreement on retention, access, and deletion. Your data is never used to train models. It's never shared with other operators. It's never sold. If the engagement ends, you get a clean handoff or a clean deletion — your call, in writing.

One high-leverage workflow, built and live in two to three weeks. Usually guest-message handling or review responses — the two highest-recurrence, highest-tone-risk surfaces in an STR operation. If we get on the call and the honest answer is "you don't need me — here's the off-the-shelf tool that does this," I'll say that. The goal is leverage in your operation, not maximizing my invoice. Operators who start small often come back six months later to extend the system, because by then they know exactly what the next bottleneck is — and they've earned the right to spend on it.

Most engagements run three to four weeks from kickoff to a working system in production. Builds run in public — you see drafts as they ship, not at the end of the engagement. Training to run the system confidently is part of the build, not a separate phase. Week one is discovery and architecture. Weeks two and three are implementation against your real workflows. Week four is parallel-run alongside your existing process so you can see the agent's drafts without committing to its outputs. By the end of week four, the system is live, you know how to operate it, and the handoff doc is in your inbox.