How we work

What we build
and what it costs.

No vague "enterprise pricing." No proposals that take three weeks. We scope fast, give honest estimates, and build things you own outright — not a SaaS subscription you can't exit.

We don't sell subscriptions, lock you into proprietary platforms, or charge for features you didn't ask for. Everything we build is yours — self-hostable, open for inspection, documented. If you decide to take it in-house later, you can. That's by design.

What we deliver

Core Engagements

Three structured ways to work with us, from a free scoping call to a full production deployment with ongoing support.

Data Readiness Audit

Start here

Before building anything, we need to understand what you actually have. The audit covers your existing data sources, how they're structured, where the friction points are, and what's technically feasible in your timeline and budget. You get a plain-language report — no jargon, no padding.

Most clients find that one or two high-leverage automations will solve 80% of the pain. We identify those and give you a specific, scoped proposal for the next step — or we tell you honestly that AI automation isn't the right tool yet.

  • Inventory of your data sources, formats, and systems
  • Assessment of ingestion complexity (PDFs, emails, databases, APIs, spreadsheets)
  • Identification of highest-leverage automation targets
  • Plain-language written report with findings and recommendations
  • Specific proposal for a pilot deployment (if applicable)
  • One round of follow-up questions included
01

Discovery Call (30 min)

You walk us through your operation — what data you have, what's costing time, what decisions are slow. No preparation needed.

02

Sample Data Review (async)

You share representative samples — a few documents, an export, whatever's typical. We assess format, quality, and ingestion path.

03

Written Findings + Proposal (within 5 days)

A plain-language document with what we found, what's feasible, and a specific scope for the next step with an honest cost range.

Engagement Fee Free Initial fit call is always free. Audit fee applied as credit if you proceed to deployment.
  • 30-minute discovery call
  • Sample data review
  • Written findings report
  • Scoped proposal with cost range
  • No obligation to proceed
Book the Call

Pilot Deployment

Most common starting point

A contained first deployment scoped to one workflow or use case — not a full enterprise rollout, not a prototype that never ships. A working system in production in 2–6 weeks depending on complexity, with real data and a real user or two.

The pilot is designed so you can assess real ROI before committing to a broader deployment. It's also the fastest way to prove to stakeholders (or yourself) that this is worth continuing. Every pilot is scoped to stand alone — it doesn't create dependencies you can't exit.

  • One contained workflow automated end-to-end
  • Document ingestion + search, or structured data pipeline, or AI-assisted output generation
  • Deployment on your infrastructure or ours (Docker + Caddy — portable)
  • Source code delivered — not a black box
  • 30 days of support and iteration included post-launch
  • Architecture documented for future extension

Examples: A document Q&A system for your legal or compliance team. A nutrition/product pipeline from ingredient input to live e-commerce listing. A job intake and tracking system for a manufacturing operation.

Starting From $3,500 Final price set at scoping. Typical range: $3.5K–$12K depending on data complexity, number of integrations, and timeline.
  • Full scoping call + written spec
  • Working production deployment
  • Source code — yours to keep
  • 30-day post-launch support
  • Architecture handoff docs
Scope a Pilot

Timeline: 2–6 weeks to production

Full Platform Build

Multi-workflow · Multi-user

A full-scale deployment covering multiple workflows, multiple user roles, and integration with your existing systems. This is the engagement type behind PleaBrain and the Nutrifunction/BFM pipeline — production platforms, not side tools.

Typically follows a successful pilot. We build on the foundation established in the pilot phase, extend to additional workflows, and deliver a platform that the team actually runs on. Includes portal interfaces, admin tools, and operational dashboards where needed.

  • Multiple integrated workflows in a single platform
  • Multi-user access with role separation
  • Custom portal interfaces for client-facing or internal use
  • Full system integrations (e-commerce, CRM, ERP, document systems)
  • Real-time and batch data ingestion paths
  • 90 days of post-launch support and iteration
  • Full documentation and architecture handoff
Starting From $12,000 Scoped after a successful pilot or detailed discovery. Typical range: $12K–$40K+ depending on scope. Multi-phase payment structures available.
  • Full multi-workflow platform
  • Portal UI (where applicable)
  • System integrations included
  • 90-day post-launch support
  • Complete documentation handoff
Start with a Discovery Call

Timeline: 4–16 weeks depending on scope

After delivery

Ongoing Support & Iteration

Optional monthly retainers for post-deployment support, iteration, and model/integration updates as the underlying AI landscape changes.

Monthly Support Retainer

Optional after delivery

AI infrastructure requires maintenance. Models change. APIs update. Clients want new features. A monthly retainer gives you a reserved block of hours to use however makes sense — bug fixes, model upgrades, new integrations, feature additions, or just questions.

This is intentionally flexible. Some months you need nothing. Some months you want to add a new workflow. The retainer rolls up to cover a backlog if you don't use it in a given month.

Essentials
$600 /mo
4 hrs/month — monitoring, minor fixes, model updates
Active
$1,400 /mo
10 hrs/month — iteration, new features, integrations
Partner
$2,800 /mo
24 hrs/month — ongoing development, priority response
Retainer Options $600+ Month-to-month. No annual commitments. Cancel anytime. Hours roll up for up to 60 days.
  • Monthly check-in call included
  • Priority response time
  • Model and dependency updates
  • Rollover hours (up to 60 days)
  • No lock-in — cancel anytime
Discuss a Retainer
Where we've already built

Four Industries, One Infrastructure

The same stack deploys across industries. Below is what "deployment" looks like in each vertical — the patterns that repeat regardless of domain.

Common questions

What people usually ask us

Why not just use an off-the-shelf AI tool?

Off-the-shelf tools are great for generic use cases. They're not designed for your specific data schema, your compliance requirements, your proprietary document formats, or your existing systems. When the thing you need doesn't exist as a SaaS product — or when data ownership matters — you need something built for your context. That's what we do.

How long does a deployment actually take?

A contained pilot — one workflow, clean data, clear requirements — can be in production in 2–3 weeks. More complex deployments (multi-source data, multiple integrations, portal UI) take 4–12 weeks. We don't pad timelines to manage expectations — if it takes two weeks, we'll tell you two weeks and deliver in two weeks.

Do we own the code?

Yes, outright. Everything we build for you is yours — source code, database schema, deployment scripts, documentation. There's no licensing fee, no platform lock-in, no "proprietary API" we control. You can take it in-house, have another developer extend it, or run it on your own infrastructure. We deliver the whole thing.

What if our data is a mess?

That's the most common situation. Real business data is almost always messy — inconsistent formats, missing fields, legacy systems with no export, PDFs that weren't designed to be parsed. The audit phase exists precisely to assess this. We've ingested everything from court documents and medical lab PDFs to manufacturer spreadsheets and email threads. Messy data is solvable, it just affects scope and timeline.

Can you integrate with our existing systems?

If it has an API, yes. We've built integrations with Shopify, Firebase, Firestore, PostgreSQL, MeiliSearch, USDA FoodData Central, OpenAI, and a range of document formats. If your system has a data export or an API we can call, we can integrate with it. If it doesn't, we assess what the extraction path looks like.

We're not a tech company. Is this still for us?

Most of our clients are not tech companies. Nutrifunction is a food delivery service. Blended Functional Medicine is a wellness practice. The law firm running PleaBrain does not have an engineering team. The value of custom AI infrastructure is highest in non-tech businesses where the competitive moat is operational excellence, not software — and where competitors are still doing things manually.

The process

From call to production

Every engagement starts free and stays honest. No retainers before we've proven value. No scope creep after we've agreed on spec.

1

Free Fit Call

20 minutes. You describe your situation. We tell you honestly if we're a fit.

Free
2

Data Audit

We review your actual data. You get a written report and a specific proposal.

Scoped
3

Pilot Build

One contained workflow in production. 2–6 weeks. You see real results.

$3.5K+
4

Full Platform

Optional. Scale to additional workflows based on pilot ROI.

$12K+
5

Ongoing Support

Optional monthly retainer. Iterate, maintain, expand as needed.

$600+/mo

Ready to start?

The fit call is free, takes 20 minutes, and you'll leave with a clear picture of what automation looks like for your specific situation — whether we build it or not.