Pricing
Priced like a utility, audited like a law firm.
AI has made software dramatically cheaper to build, but only if someone manages the meter honestly. We charge a predictable retainer for senior human judgment, pass through the AI build costs with a stated margin, and show you every credit. Here's exactly how that works.
Fixed bands, not blank checks
Every engagement starts with a scoped price band. You know the ceiling before a line of code is written, and we inherit the risk of getting the estimate wrong.
Every credit, audited
AI build tools meter usage in credits. You get a full, per-project audit trail of exactly what was consumed, the same ledger we see. Pass-through cost plus a stated margin, never buried.
We carry the experimentation risk
Rapid prototyping burns tokens by design. That cost of exploration is ours, not a surprise line item on your invoice. You pay for outcomes, not our learning curve.
Rates that adapt honestly
If underlying AI platform pricing moves, pass-through rates adjust at renewal, with notice and the same audit trail. Your margin of predictability stays intact.
The anatomy of an invoice
Two lines. No fog.
Human time
Senior product and engineering judgment: strategy, planning, architecture, review, and accountability. Build work is scoped per phase and priced as a band. Guidance and planning is billed hourly, like a fractional PM on retainer. Either way, the number is agreed before work starts.
AI platform credits, passed through
What the AI platforms actually consumed on your project, at cost plus a stated margin on build work, with a per-project audit trail you can inspect line by line. Efficiency is our job; visibility is your guarantee.
Two ways to work with us
Build
Retainer + metered credits
We act as your fractional product and dev team, directing a virtual crew of AI builders from idea to shipped product. You pay the engagement retainer for our judgment and accountability, plus AI build credits at a stated pass-through rate with full audit visibility. We absorb the cost of prototyping and experimentation; the band we agree on is the number you pay.
Guidance
Hourly, product-planning focused
For teams who want product direction without a build engagement. We run planning and scoping sessions in the AI platform on your behalf, take its raw planning output, and shape it into specs, roadmaps, and decision docs using a broader AI toolkit before it reaches you. Billed hourly, because the value here is judgment, not tokens. Any platform credits this work consumes are passed through at cost, itemized: it is usually a small number, and showing it to you is the point.
Calculator
Run the numbers yourself
Pick an engagement type, set the dials to something that looks like your project, and see how the invoice breaks down.
Estimate your month
The retainer is a balance you draw against. Move the sliders; every number is one you can question on a call.
drawn from a $4,000 monthly retainer
- Meetings (2 scheduled − 2 included = 0 × $150)the retainer includes a standing meeting allowance; extras carry a fee component
- $0
- AI credits (400 @ $1.30)
- $520
- Hosting & third-partyat cost, or billed to accounts you own — see the Hosting tab
- $225
- Total drawn this month
- $745
- Kickoffsigned scope and agreed milestone plan$020%
- Milestone 1working demo reviewed and accepted by you$015%
- Milestone 2working demo reviewed and accepted by you$015%
- Milestone 3working demo reviewed and accepted by you$015%
- Milestone 4working demo reviewed and accepted by you$015%
- Acceptancefinal handover: code, data, and accounts in your name$020%
- Most you've paid before working software is in your handsyour downside if we disappoint
- $0
- Work we've funded before being paid for itour downside if you walk away
- $0
- Formal go / no-go checkpointschances to redirect, cut, or add before the money is spent
- 5
The balanced default, and what most of our work uses. Risk is split into small slices, and each milestone is a real demo you accept before it's billed — so feasibility, usability, and value risk get retired early, while you keep the option to redirect, cut, or add scope at every checkpoint.
Projects sit outside the retainer and follow discovery-then-build: a small, paid discovery phase produces the brief and architecture, then the build price is negotiated and fixed before work starts. All three options land on the same outcome — the same scope, delivered on the same date, for the same total. They differ only in who carries the cash risk in the middle, and in how many times you get to steer. If scope changes mid-flight, it's repriced in writing (a signed change order) before work continues, so the number you approved is the number you pay.
Illustrative only. You always see the actual ledger — every meeting, credit, and pass-through charge against your retainer balance. The balance reloads on the 1st of each month, and if it runs low mid-month we agree a top-up (or pause) before any work outruns it — never a surprise invoice after the fact.
Which one is right for us?
Ongoing product & development
You have a steady stream of ideas, fixes, and features — and want a team on call. A monthly balance you draw against, reloaded on the 1st.
Typical: $1,500–$8,000/mo
One defined build, one locked price
You know exactly what you need built. After a short paid discovery, the price is fixed — payable upfront, per milestone, or at completion.
Typical: $10k–$75k per project
Advice, not hands-on building
You have your own builders (or build in Lovable yourself) and want a senior product brain a few hours a month. Hourly, no commitment.
Typical: $300–$3,000/mo
Not sure? Most clients start with a fixed-fee project and roll into a retainer once the ideas outpace the original build. A 30-minute call settles it — there's no wrong door.
Hosting & ownership
You own the software. Always.
Two ways to keep your product running after launch. Both end with you holding the keys: no hostage hosting, no proprietary lock-in, no rebuild tax if you leave.
Managed by us
We run your app on the AI platform's managed cloud: database, authentication, storage, functions, and hosting in one place. You get a monthly care plan that covers uptime monitoring, updates, backups, and a set amount of change work. Runtime costs are itemized at cost.
- ▸Fastest to launch, nothing for you to operate
- ▸One monthly number: care plan plus runtime at cost
- ▸We hold the pager, you hold the roadmap
- ▸Export to your own infrastructure whenever you want
Full source control and ownership
Take the entire codebase into your own repository and run it on standard infrastructure: Vercel or Netlify for the app, Supabase for database and authentication, your own domain and email. Accounts are in your name, billed directly to you, with no markup passing through us.
- ▸Your repository, your cloud accounts, your bills
- ▸Standard tooling any developer can pick up
- ▸We hand off with docs, environment setup, and a walkthrough
- ▸Keep us on retainer for changes, or don't
What ongoing cost actually looks like
A handful of users, light data. Runtime typically lands in the low tens of dollars a month.
Real traffic, files, and email. Runtime usually sits in the low hundreds, scaling with usage.
Optional. A care plan for steady improvements, or per-request work billed the same transparent way.
We set usage alerts on every project, so nobody discovers a cost surprise on an invoice.
Fair questions
Asked by every smart client
Why not just charge one flat fee?+
Flat fees force a bad trade: either we pad for worst-case usage (you overpay) or we absorb unbounded AI costs (we cut corners). Metered pass-through with a fixed band keeps the price honest and the incentives aligned: we stay efficient because waste is visible to both of us.
What exactly am I paying for?+
Two things, stated separately: human time (strategy, planning, architecture, review, accountability) and the metered AI platform credits your project actually consumes, passed through with a clear markup on build work. Build engagements pair a fixed band with metered credits; guidance and planning is billed hourly. No blended mystery number.
How does guidance and planning differ from a build engagement?+
Guidance is advisory: we direct planning inside the AI platform, take its raw planning output, and edit it heavily into specs, roadmaps, and decision docs with other AI tools before presenting it to you. It is billed hourly, like a fractional product manager. Build engagements go further: we direct the AI builders through to shipped software, priced as a fixed band plus metered credits.
Why pass through credits on guidance work at all?+
Because the ledger is small on advisory work, and hiding it would undermine everything else we claim. Guidance sessions consume a handful of platform credits; passing them through at cost, itemized, keeps the audit trail complete and shows you where the real value sits: in the judgment and the edited work product, not the tokens.
What about third-party and hosting costs?+
Anything your product depends on (hosting, email delivery, maps, payments, data providers) is itemized up front in your plan. Where possible we set services up in accounts you own, so you're never locked into us and there's no markup fog.
What happens if AI platform prices go up?+
Pass-through rates can change at renewal with advance notice, never mid-engagement, never retroactively. The fixed band for your active scope holds. This is the honest version of pricing in a market where token costs are still finding their level.
Can I see usage during the build?+
Yes. The credit ledger is available per project, and we review it together at each weekly checkpoint. If a scope decision is about to burn unusual credits, you hear about it before it happens, with options.
Who owns the code, and can I host it myself?+
You own it, in every case. You can leave the app on the managed cloud and let us operate it, or take the full source into your own repository and run it on standard infrastructure like Vercel for the app and Supabase for the database and login. We hand off with documentation, environment setup, and a live walkthrough. Nothing about the build is proprietary to us.
What does hosting cost each month?+
It depends on real usage, not a made-up tier. A small internal tool usually runs in the low tens of dollars a month; a customer-facing product with real traffic and email typically sits in the low hundreds. If we manage it, runtime is itemized at cost alongside a care plan. If you host it yourself, the accounts are in your name and you pay the providers directly with nothing flowing through us.
Do you still do fixed-price projects?+
Yes. Most engagements are fixed bands by phase (strategy, build, launch). The metered model governs the AI consumption inside those bands, not a blank hourly meter. Predictable total, transparent internals.
Want a band for your project?
One 30-minute call usually gets you a real number.