How we work

We embed with your team, find the workflow costing you the most, and build the system that runs it, starting from our own products where they fit. You keep the code, the data and the accounts.

The forward-deployed engineer, at small-business scale

The largest AI companies send engineers into their customers’ operations instead of shipping software over the wall. We work the same way, for organizations that were never offered it.

Where the term comes from

“Forward-deployed engineer” is Palantir’s term. Instead of selling software and leaving the customer to make it work, Palantir stationed engineers inside the customer’s operation to learn the real workflow, build against it, and stay accountable for the result. OpenAI and Anthropic now send forward-deployed engineers to their largest customers.

The reason is the last mile. What separates a capable product from a working system is knowledge of how one particular organization runs: who touches what, where the exceptions are, which step lives only in one person’s head. You learn that by sitting where the work happens.

Why small organizations never got one

An engineer embedded in your operation for weeks was always an enterprise line item, so nobody offered it to a business with six employees. Small organizations got off-the-shelf software built around someone else’s workflow, and the permanent job of bending their own work to fit it.

What changed

AI multiplies an engineer’s time. One senior engineer working with AI now delivers what recently took a team, which puts the embedded model within a small organization’s reach.

The expensive part was never the code. It is the judgment about what to build and what not to, and that is what you pay for.

Code comes after the thinking

Requests usually arrive small: we need a new website, or we collect everything in Google Forms. Once we have watched the work, the need is often larger and clearer, such as a public site, a staff dashboard, one contact database, role-based workflows, bilingual communication, automated follow-up and reporting, on a foundation you own.

So we don’t hand you a strategy deck. We build the system, and it keeps running after we leave.

The full argument is a sixteen-slide talk: What AI actually means for a business like yours

How an engagement runs

Five steps from the first conversation to a system that runs on your own accounts. You see a fixed quote before anything is built.

  1. A conversation

    Tell us where the work strains. No jargon, no obligation.

  2. We embed

    A few days inside your actual operation, watching how work really happens.

  3. The blueprint

    The workflow, redesigned on paper, with a fixed quote attached.

  4. The build

    A working prototype in weeks. You steer while it takes shape.

  5. Handoff and care

    It runs on your accounts, documented, and keeps running after we leave.

When one of our products fits, the build starts from it instead of a blank page. Our products

What the practice brings

Four disciplines that rarely sit in one practice, so the same people who understand your operation also design and build for it.

Senior software engineering

Decades of hands-on experience building real systems, not only prototypes. AI writes code quickly, but speed without architecture makes fragile systems. Experience decides security, maintainability, accessibility and how the system is deployed, the parts quickly generated projects tend to skip.

Solution architecture for AI-era work

Turning an unclear situation into workflows, data structures, dashboards, automation and systems you can deploy. That starts with the people: who needs to see what, where information leaks or gets duplicated, and what should be automated or stay human.

Design and communication

A long background in graphic design, visual communication, branding and layout. What your staff and the public see is designed for them, with no templates.

Humanistic, strategic thinking

Experience as a writer, with published books, and a long interest in how people, organizations and technology meet. It shapes the question each engagement starts from: if you were starting today, how should this work?

Who it is for

Organizations that know they need to change but don’t yet know what the right system looks like.

Nonprofits

Program management, volunteer coordination, reporting, and the records that earn public trust.

Education and youth organizations

Assessments, enrollment, parent communication, certificates and bilingual experiences, with outcomes you can measure.

Community networks

Member portals, partner management, event workflows and dashboards.

Small businesses and founders

A business that has outgrown its static website, or a leader with a clear idea who needs someone to turn it into a working platform.

Quoted by the work involved

These ranges are typical. Every quote is agreed before building starts, and the blueprint is where the numbers become exact.

Workflow blueprint

$500 to $1,500, fixed

We embed, map one workflow end to end, and hand you the redesigned operating model. Yours to keep, whoever builds it.

Platform build

$2,000 to $8,000, typical

The blueprint, built: a prototype in weeks, then a production system on accounts you own, starting from our products where they fit.

Changes on demand

$200 to $2,000 per change

A new report, a new workflow, a redesign. You pay when you need something. There’s no monthly retainer.

When one of our products fits, it is scoped and quoted as part of the build. Pricing scales to the size of your organization, not to enterprise budgets. If yours is small, say so.

Start with one workflow

Tell us how the part of your operation that annoys you most works today, and we’ll tell you what it should become before any talk of a contract.