Why don't you publish prices?

Because a number without a scope is a guess, and the honest range is wide enough to be meaningless. A single follow-up agent over a clean CRM and a five-agent deployment with a voice line on your main number are different orders of magnitude. What we can tell you before a call is exactly what moves the number — that's the section above — and what we will tell you on the call is a real figure for your actual situation, not a range.

How does the money actually work across the sprint?

Two components. A one-time build fee for the 90-day sprint, invoiced in staged milestones tied to the phases rather than all up front. Then a monthly operating retainer that starts when the build goes into operation. Third-party costs — model usage, telephony minutes, SaaS seats — sit in your own accounts and bill to you directly, so you keep the vendor relationship and see exactly what they cost.

What's the smallest sensible engagement?

One agent, one clear problem, where the cost of the thing going unhandled is obvious. Usually that's inbound: the calls, forms, and messages that get missed. It's a smaller build, it goes live faster, and it gives you something real to judge us on before you commit to more. We'd rather start there and earn the next agent than sell you five and hope.

Am I locked in after the sprint?

No, and never. The retainer is month-to-month with thirty days' notice and no penalty. There is no annual contract and no minimum term. If you leave you keep every outcome the systems produced, all of your data, and the documentation and runbooks for how your operation works. We would rather keep earning it every month than rely on an agreement that traps you.

What if the ROI doesn't show up?

Then we've got a problem to solve together, and staged go-lives mean we find out early rather than on day 90. The first system is live in week five or six precisely so there's a real number to look at while there's still time to change course. If a specific automation isn't returning what we projected, we say so in the weekly review, and we either fix it or we stop doing it. If we thought it wouldn't pay off, we should have told you before the build — and that's a call we do make.

Do you work with businesses smaller than your case studies?

Yes, and most of our engagements are. The production systems we publish came out of enterprise and healthcare operations, and that's stated plainly wherever they appear. What carries over is the architecture and the standard, sized to your business. A ten-person service company gets the same build standard as a two-thousand-person operation; it gets fewer agents.

The engagement model

What you're buying, and on what terms.

Two components, month-to-month after the build, and no dollar figure on this page — because a price without a scope is a guess. What is on this page: exactly what moves the number, what's included, what isn't, who owns what, and who this is wrong for.

If you want the week-by-week of how the build runs, that's a different question and it lives on its own page.

What sets the price How the 90 days run

Structure: One-time build + monthly operating retainer
Build fee: Staged by milestone, not up front
Retainer term: Month-to-month. 30 days' notice
Minimum term: None. No annual contract

01The shape of it

A build, then an operation. Priced separately, on purpose.

Most AI vendors sell you one or the other: a project that ends and leaves you holding it, or a subscription to software you have to staff. The two-part structure exists because neither of those works — a system nobody operates degrades, and software you have to run yourself isn't leverage, it's another job.

Component one · one-time

The 90-Day Sprint

Discovery, architecture, and the full build of every system we agree on — deployed inside your business, integrated into the tools you already run, documented as we go. Fixed scope and fixed timeline, both agreed in writing before kickoff. Invoiced in milestones tied to the phases rather than demanded up front.

Ends with

Your AI team live in production, documented, with a runbook and a kill switch per agent.

Component two · monthly

The operating retainer

We run what we built. Monitoring and alerting wired to us, agents retuned as models and your business change, issues fixed before you notice them, your dedicated Specialist and your direct channel, and new systems built on request. Month-to-month, thirty days' notice, no minimum term.

Exists because

Unoperated AI systems degrade. Month six should beat day 90, and that only happens if someone is accountable for it.

Third-party costs sit with you. Model API usage, telephony minutes, and any SaaS seats bill to your own accounts, sized during architecture so there are no surprises. We don't mark them up and we don't hold the vendor relationship — that's yours.

02What sets the price

Five things move the number. Here they are.

Every firm in this industry says pricing is custom, which tells you nothing. These are the actual variables — read them and you'll know roughly where your build sits before you speak to anyone.

01
How many agents, and how hard each one is
A follow-up agent working over your existing CRM is a different build from a voice agent that has to hold a real conversation on your main line at sub-second latency. Scope follows difficulty, not a tier chart.
02
The state of the systems we're connecting to
A clean CRM with a real API is fast. A decade of records in spreadsheets, or a system of record with no integration surface, means data work before agent work — and we'd rather price that honestly than discover it in week 6.
03
How much sits behind a human checkpoint
Regulated work, anything that moves money, and anything that signs your name gets an approval step. More checkpoints means more workflow design, and it is usually the right call.
04
Volume and what happens at peak
Two hundred calls a day and twenty thousand are different engineering problems. We size for your peak, not your average, because the busiest hour is the one that costs you customers.
05
How much we run after launch
The retainer scales with what we're operating and how much we keep building. A stable four-agent deployment costs less to run than one where you're shipping something new every month.

Why no numbers on this page

An honest range across everything above is wide enough to be useless, and a narrow one would only be true for the client it was written for. So we don't print a figure we'd have to walk back. On the call you get a real number for your actual scope, with the reasoning attached — and if the scope is smaller than you feared, we say that too.

Model it yourself first

The ROI calculator works the other end of the equation: what these systems return in a business your size.

Run your numbers

03Scope

What's in it — and what genuinely isn't.

The exclusions matter more than the inclusions. Anything a vendor leaves unsaid at this stage is something you find out about in week eight.

Included in every sprint

Discovery, architecture, and the full build of every agreed system
All integration work into your existing stack
A dedicated AI Specialist who knows your business by name
A private messaging channel — a direct line to our team, not a ticket queue
Weekly progress reviews and a live dashboard you can open any time
Documentation and a runbook for every agent we ship
Team training on every system that touches your people
Staging and production environments, so nothing untested reaches your customers

Not included

Third-party software and usage costs
Model API usage, telephony minutes, and any SaaS seats sit in your accounts and bill to you directly. We size them during architecture so there are no surprises, and you keep the vendor relationship.
Rebuilding your system of record
We work with the CRM you have. If your business genuinely needs to migrate platforms, that's a real project and we'll say so plainly rather than hiding it inside a sprint.
Work we don't think will pay off
If an automation you're asking for won't return more than it costs, we'll show you the math and recommend against it. That has happened, and it will happen again.

04Ownership & IP

You own every outcome. We own the upkeep.

Echelon builds and operates the AI systems we deploy for you. You license access to those systems for as long as we work together, and you retain full ownership of every outcome, result, and work product they generate — every lead answered, every quote sent, every dollar collected. We run the engine; everything it produces is yours.

Yours outright: Every result the systems produce — leads, bookings, quotes, collections, transcripts, and the customer data behind them. Your data stays in your accounts.
Yours to take: The documentation, runbooks, and process maps we write for your build. If you leave, you leave knowing exactly how your operation works.
Licensed to you: The agent systems and the infrastructure they run on. We maintain, monitor, and improve them — that is what the retainer buys, and why nothing rots after launch.
Never ours to sell: Your data does not train our models or anyone else's, and it is never pooled with another client's. Per-client isolation is a build standard, not an upgrade.

Said plainly because vagueness here is what buyers are right to distrust: we are not handing you a codebase to maintain, and we are not holding your business hostage either. You own the results and your data; we own keeping the machine running.

05Terms

The commercial terms, in plain language.

No defined terms, no cross-references, nothing that needs a lawyer to parse. If any of this reads differently in the agreement we send you, the agreement is wrong.

Standard engagement terms

Sprint term: Ninety days, fixed scope and fixed timeline, agreed in writing before kickoff.
After the sprint: A monthly operating retainer, month-to-month. No annual contract and no minimum term.
Notice to leave: Thirty days, any time, no penalty. We'd rather you stay because the systems earn their keep than because an agreement traps you.
If you leave: You keep every outcome the systems produced and all your data, and we hand over the documentation and runbooks so your team knows exactly how the operation works.
Scope changes: New systems mid-sprint get quoted separately rather than silently absorbed — that's how a 90-day build turns into a 12-month one.
If something breaks: Alerting is wired to us, so most issues are fixed before you notice. If a failure on our end cost you real money, we make it right.

Our commitments, in writing

Staged go-lives — your first system is working weeks before day 90
You own every outcome the systems produce
Month-to-month after the sprint. No lock-in, no minimum term, ever
Built in your accounts. Your data never trains anyone's models
Senior architects on your build — the same people behind the production numbers we publish
If we're not the right move for your business, we tell you on the first call

06Fit

This is a bad fit for some businesses. Better you know now.

We qualify hard on the first call, so we'd rather you self-qualify before it. If you're in the right column, the call will be worth your thirty minutes. If you're in the wrong one, we'll say so on the call anyway.

Worth a call

You have real inbound volume — calls, forms, or messages — and you know some of it goes unanswered
There's a repeatable process a person is running by hand today
One person can make the decision and stay in the loop for an hour a week
You want the thing operated, not handed to you as software to staff

Not yet, or not us

You want a demo to show a board, with no intention of running it in production
Nobody on your side can commit an hour a week — the build needs a counterpart
The underlying process is genuinely broken; automating it would just make the mess faster
You need it live in three weeks. We'd be setting you up to fail, and we'll say so

07Getting to a number

Call, proposal, agreement. No pressure in any of the three.

How you actually get from this page to a figure. Nothing here is a sales sequence with a deadline attached.

Step 01 · 30 minutes, no cost

The strategy call

We walk your operation: where work comes in, where it stalls, what your team spends its hours on. By the end of the call you know which agent we'd build first, roughly what it's worth in your business, and whether we think this is worth doing at all.

You leave with something useful either way. If we don't think we should work together, you get that answer and the reasoning, on the call, for free.

Step 02 · within a few days

The written proposal

A scoped proposal: the specific systems, the order we'd build them in, the outcome each one owns, the build fee broken into milestones, and the monthly retainer. Third-party costs estimated separately so you can see what's ours and what's a vendor's.

One number, one scope, no tiers to compare and no artificial urgency. It stays valid while you think about it.

Step 03 · when you're ready

Agreement and kickoff

A statement of work matching the proposal exactly — the phases, the deliverables, the terms above. Signed, kickoff gets booked, and your channel opens the same day.

Fixed scope and fixed timeline are in the document, not just on a website. Anything new mid-sprint is quoted separately rather than absorbed, which is what keeps 90 days from becoming twelve months.

Want the build detail first? The phases, the deliverables, the cadence, and an honest account of what goes wrong all live on the process page.

See the 90 days week by week

08Commercial questions

What people ask before they send the agreement back.

Engagement & pricing

Get a real number

Thirty minutes gets you a scope and a figure.

Not a range, not a brochure. We walk your operation, tell you which agent earns its keep first, and give you an actual number for your actual business — or tell you plainly that it isn't worth doing yet.

Book a Strategy Call How the build runs

Prefer to talk now? Call or text (805) 285-2713

No commitment
Custom scope on the call
Month-to-month after the build