This is for you if you run a business where people ask about availability and price before they buy.
Event and party venues. Boutique gyms and fitness studios. Catering and events services. Creative and experience studios. Specialty trades that quote before they work.
Five to fifty people. Somewhere between fifteen and a few hundred new inquiries a week, arriving from more than one place. An owner who can make a decision without a committee.
We work in New York, in Puerto Rico, and remotely — in English or Spanish.
The inquiries arrive. That was never the hard part.
They come from the website form, and email, and Instagram, and someone calls, and someone messages on WhatsApp. There is no single line they all land in, and no one person whose actual job is answering them. So they get handled by whoever is not on the floor right now.
Which means the busier you are, the worse you answer. And busy is exactly when the inquiries come.
The person who can quote accurately is the same person delivering the service. The pricing lives in someone's head. The good proposals — the ones that actually won the job — live in someone's sent folder. Every reply gets written from scratch, by someone doing three other things.
Some of them go cold. You know it is not zero.
Most owners we speak to cannot tell us what percentage of their inquiries never got a real answer. That is the whole problem in one sentence. You cannot fix what you cannot see, and right now nobody is counting.
What another year of this costs
The bookings you never hear about again
Not the ones you lost on price. The ones where someone reached out, waited two days, and booked the place that answered in twenty minutes. You will never know their names.
Your evenings
You are answering messages at eleven at night and on Sunday, because that is when there is time. It has stopped feeling temporary.
Your team's confidence
The people who are excellent with a customer standing in front of them feel like they are failing at the inbox. They dread it. That is not a training problem — it is a structural one, and no amount of trying harder resolves it.
Your ceiling
The business cannot grow past your personal attention, because the thing that makes it work does not survive you being unavailable. A second location is not a real option while this is true.
The knowledge walking around loose
The pricing logic, the way you handle "can you do better on that," the tone that wins the job — it exists in one or two heads. When those people leave, it leaves.
The version where this is handled
Every inquiry — whatever channel it arrives on — lands in one place with someone's name on it.
Within minutes, there is a draft reply waiting: correctly priced, in your voice, referencing the right package, ready for a person to read, adjust and send. Your team stops writing from scratch and starts reviewing. That is a completely different job, and a much faster one.
Before any of that gets built, somebody has to work out whether it is actually worth building for your business. That is what these three weeks are.
- Nothing reaches a customer without one of your people approving it. Not once, not ever. The assistant drafts; a human decides.
- At the end of every month you can see the numbers: how many came in, how fast you answered, how many turned into bookings. Not a feeling. A count.
- The pricing logic and the winning tone live in a system your team owns — not in one person's head.
Three weeks, ten conversations, one decision
We find out how it really works
A kickoff call, then forty-five minutes each with three to five of your people. Not the org chart version — the real one, including the workarounds nobody has mentioned to you. We look at your actual inquiry traffic and start counting: how many arrive, through which channels, how long they wait.
If you have never counted before, that is normal. We bring a tally sheet and start on day one.
We map it and put numbers on it
Every routine workflow we find goes on a list, ranked by how much time is recoverable. The top three get properly costed. We score your readiness across six dimensions — your data, your tools, your people, your process, your oversight, your risk — and name the specific gaps rather than giving you a grade.
Midway through, we show you the picture so far. If we have misunderstood something, this is where you tell us, not at the end.
We design one thing and show it working
We take the single workflow with the most recoverable value and design it properly: what the assistant drafts, what your person reviews, where it stops and asks for help, what happens when it gets something wrong.
Then we build a working version on made-up data, in your industry's vocabulary, and show it to you. Not a slide. A thing that runs.
Sixty minutes with you and anyone else who would need to approve a pilot. What we found, what it is costing you, what to do first, what it would cost, what could go wrong — and our honest recommendation, including "not yet" if that is the truthful answer.
Nine things, all of them yours to keep
A measured baseline
How your inquiry workflow performs today, in numbers rather than impressions.
A ranked opportunity list
Every routine workflow we found, ordered by recoverable value, top three costed.
A readiness score
Six dimensions, with the specific gaps named — not a grade.
Current and future workflow maps
How the work moves today, and how it would move instead.
One workflow designed in full
With a named person approving every step that reaches a customer.
A live demonstration
A working version of the proposed assistant, running on made-up data in your vocabulary.
A risk register
What could go wrong, how likely it is, and what would prevent it.
A costed pilot recommendation
What it would cost, how long, and what result would justify continuing.
An executive summary
Six to ten pages, plus the readout deck, written for someone who was not in the interviews.
Your time: about eight and a half hours across three weeks. Roughly three and a half of them are yours personally.
Questions you cannot answer today
Everything we produce is yours. The maps, the numbers, the workflow design, the risk register. If you take that document to a different firm, or hand it to someone on your team, or put it in a drawer for six months — all fine. It is not written to be useful only to us.
- How many inquiries you actually receive each week, and through which channels
- How long, on average, someone waits for a real reply
- How many never get one
- Which single workflow in your business is costing you the most recoverable time
- What it would cost to fix it — including the parts most people leave out, like the time your staff would spend reviewing
- Which of your work an AI assistant should never touch, and why
- Whether this is worth doing at all, right now, for your business
Who this is for
- You take fifteen or more new inquiries a week, from more than one channel
- You can approve a $2,500 decision without a board or a committee
- You have five to fifty people, and someone on staff who is "the computer person"
- You want to know the real answer, including if the real answer is no
- You can free up three to five of your people for forty-five minutes each, in the next three weeks
Who this is not for
- You want something built, not evaluated. Fair enough — but this is diagnosis. If you already know exactly what you want built, ask us about a pilot instead.
- You handle medical records, privileged legal material, or student records. We do not work with those categories yet, and we would rather say so now than discover it in week two.
- Someone else has to approve the spend. Not a criticism — it just means the founding cohort is the wrong fit, because the whole point is moving quickly.
- Fewer than about fifteen inquiries a week. The arithmetic does not work yet. Come back when volume has grown, and we will say the same thing then for free.
- You are hoping to reduce headcount. That is not what this is for, we will not design toward it, and we would be a poor choice.
- You want a guarantee of a specific return. Nobody can honestly give you one before measuring, and we are not going to be the first to try.
What "founding client" actually means here
No euphemisms. Here is the trade.
What you get
- The full evaluation — every deliverable listed above, nothing removed
- A live demonstration built for your workflow, which is not part of the standard engagement
- The founder delivers it personally. Every interview, every map, every hour. That will not be true in a year
- Thirty days of questions afterward, by email, at no charge
- Founding pricing held for twelve months on anything that follows
What we ask for
- Forty-five minutes of honest feedback, recorded, after the readout. What was useful, what was confusing, what we missed. This is the actual reason the founding cohort exists
- Permission to ask — not to assume — about a written case study after sixty days
- Two introductions to other owners, if you were glad you did it. Not before then, and not if you weren't
- Patience with a process that is being refined. Which brings us to the part most companies leave out.
What is finished, and what isn't
That last line is the one we thought hardest about publishing. It is true, and you would find out anyway. If it is a dealbreaker, it should be — and you should wait until we have references and pay the standard price for the certainty.
Why only five
Not a marketing device. One person delivers these, three can run at once, and five is what fits in ten weeks at the quality described above. When the five are taken, this page will say so. As of October 2026, all five places are open.
$2,500
$1,250 to reserve your place. $1,250 when the readout is delivered.
The standard price for this scope of work is $4,000. Founding clients are not paying a discounted rate for less work — they are getting the full scope, plus the demonstration, at the price of our smaller tier. The difference is what the recorded feedback session is worth to us.
Three weeks from kickoff to readout. Places open in three waves — pick the one that suits your calendar, starting within ten days or as late as ten weeks out.
What is not included
Anything being built, connected or configured. Legal, tax or compliance opinions. Work involving medical, legal-privileged, or student records. A second location or a second function.
For nonprofits: the same evaluation is available, and we adjust for organization size. Ask.
How you know you got what you paid for
We are not going to offer you a money-back guarantee, and we want to explain why, because the reason matters. What you are buying is an honest finding. If "satisfaction" were the standard for getting paid, we would have a quiet incentive to deliver findings you enjoy. That is exactly the failure we are trying to avoid.
So instead of a feeling, you get a checklist. Everything under "What you get" either exists at the end of three weeks or it does not. It is verifiable by you, not arguable. If any of it is missing, you do not pay the second half.
- If we do not deliver by the readout date, you get the deposit back in full, plus everything produced up to that point.
- If something in the report is wrong, we fix it. You get ten business days after the readout to tell us the inventory does not reflect your operation. Corrections are incorporated within five business days, at no charge.
And the one that matters most: if the honest answer is that AI is not the right move for you right now, we will say so in writing — and you will still have every map, number and document. That is written into the agreement, not just onto this page.
Asked before, answered straight
"Can't I just do this myself with ChatGPT?"
For a single reply, yes — and if you have not tried, you should before you talk to us. The difference is between a good answer once and the same good answer four hundred times, from a system that reads your actual price list, follows your rules about deposits and minimums, hands each draft to the right person, and keeps a record of what it did. Most owners we speak to have already tried the first thing. They call because they hit the wall making it repeatable.
"$2,500 to be told something? I'd rather you just built it."
Understandable, and sometimes right — if you already know exactly what to build. Most owners don't, and building the wrong workflow costs far more than three weeks. A pilot runs $6,000 to $14,000. Spending $2,500 to make sure it is aimed at the right thing is the cheaper mistake to avoid. And if the evaluation finds you don't need a pilot, that is $2,500 that saved you nine.
"Will my customers know they're talking to a robot?"
They will be talking to your staff. The assistant writes a draft; one of your people reads it, changes what needs changing, and sends it. Nothing reaches a customer unapproved — that is a design rule we do not bend, and it is in every workflow we produce.
"Is this going to cost people their jobs?"
Not through anything we design. We do not build toward headcount reduction, and if that is your goal we are the wrong firm. Every workflow we design assumes the same people, doing better work, with less of it being retyping. We also tell you plainly in the readout: hours returned are capacity, not payroll savings — and we will not pretend otherwise to make the numbers look better.
"What if you find out AI won't help us?"
Then we say so, in writing, and you still have the operations map — which is worth having regardless. It has happened in this line of work and it will happen again. We would rather be the firm that says it.
"How much of my time is this?"
About eight and a half hours across three weeks, roughly three and a half of them yours personally. Everything else is your team, forty-five minutes each. We schedule all of it during the kickoff call so it does not become a scheduling problem.
"What happens to our information?"
You send redacted samples — we tell you exactly what to remove before you send anything. It lives in one folder only we can see, and it is destroyed or returned within thirty days of the readout unless you ask otherwise. We do not put your data into any system you have not approved, and we do not keep it on our own infrastructure.
"Are you certified for HIPAA / FERPA / legal confidentiality rules?"
No, and we will not claim to be. We do not currently take on work involving medical records, student education records, or privileged legal material. If your business touches those, we will tell you at the first call rather than the fourth.
"You're new. Why should we trust you with this?"
We are new, and there is no way to pretend otherwise. What we offer instead of a portfolio: a fixed price, a bounded scope, a written list of what you receive, dates you can hold us to, and a section on this page that tells you which parts of our process are established and which are still being built. If that is not enough, it should not be — and you should wait until we have references.
"What if we want to do the pilot with someone else?"
Take the documents and go. They are written to be used, not to lock you in. We would rather you built the right thing with someone else than the wrong thing with us.
Loli Mari Montalvo
Loli owns and runs Isla Metro Paint & Sip, an events business in New York City with exactly this inquiry problem — private parties, corporate bookings, questions about availability and price arriving from five directions at once. She is not a consultant who read about the inbox. She answers one every week.
She also teaches as an adjunct professor, which is most of what this work is: explaining things clearly to people who did not ask to become technical. With a Master's focused on Artificial Intelligence from Purdue and an MFA from Pratt, she brings together technology and creative practice — and she cares that the drafts sound like you, not like a machine, because that is what wins the job.
She started Sunset for Everyone because the gains of AI should be for everyone, not only for companies with an engineering department. She works in English and Spanish, in New York and Puerto Rico.
Five places. Three weeks. One honest answer.
Fourteen questions, about four minutes. We reply within two business days either way — including if we do not think you are a fit, and we will tell you why.
Rather talk to a person first?
Twenty minutes on the phone is enough to tell whether the founding cohort is the right fit. No deck, no pitch.
hello@sunsetforeveryone.org · (646) 221-5981 · New York · Puerto Rico · English or Spanish