CMCarl MannSolutions Architect
Engagement · 01

Custom web applications, end-to-end.

From the schema to the screen, written by the person who'll be there at launch.

Greenfield builds in React + Node + PostgreSQL -- internal tools, customer-facing dashboards, vertical SaaS products. Every layer written by the same engineer, so when something needs adjusting six months in, the email goes to the person who actually remembers why it was built that way.

8-14wk typicalReact + TS+ Postgres1engineer, the whole way
What you get

The whole vertical, one head.

Not a team of three contractors with overlapping calendars. One engineer who can hold the design in their head from the data model to the button click -- and the architecture notes that make it survive after handoff.

01

Schema to screen

Database design, ingestion, API, and UI -- not handed off across three vendors. One person holds the whole vertical, so the trade-offs on day forty add up instead of contradicting each other.

02

Demos every Friday

Working software in your hands every week, not a slide deck six weeks in. If the brief drifts, you see the drift the same week it happens -- before it ossifies into a rewrite.

03

Documented to survive me

Architecture notes, runbooks, and onboarding docs as deliverables -- not extras. The next developer reads the code in an afternoon, not a fortnight.

Engagement model

Fixed scope.
Fixed launch date.

The shape of the contract before the first commit -- so the project has a "done", and the calendar has a date.

  1. 01

    Brief to proposal

    A two-week conversation: requirements, constraints, the systems already in play. Output is a written proposal with scope, milestones, and a fixed launch date.

  2. 02

    Fortnightly milestones

    Billed against working demos, not hours. Each milestone is a vertical slice you can use -- not a status meeting.

  3. 03

    Friday demo, Monday note

    Demo at the end of every week, a written status note at the start of the next. No surprise overruns at month three.

  4. 04

    Launch + 30 days

    Thirty days of post-launch support included as a matter of course. Longer retainers available -- typically for the first quarter after a public launch.

Scope notes

A good fit, and a not-quite-fit.

Good fit

A bounded problem with a clear "done".

  • A tool, a portal, a dashboard, a workflow -- something with an edge.
  • A stakeholder who can attend a weekly demo and make decisions on the spot.
  • Modern stack, modern hosting, a team that can read TypeScript without flinching.
  • A "ship it and improve" temperament, not a perfectionist holdup until paragraph 9.b of the spec.

Less of a fit

A bet that needs a co-founder, not a contract.

  • A platform-defining moonshot for an unfunded idea -- those need equity and time, not a fixed-scope engagement.
  • Native iOS or Android. This is a web practice -- happy to recommend specialists.
  • On-prem deployments behind a procurement gauntlet. Worth a conversation, but not the typical shape.
  • "Just make it look like ChatGPT but for X." Reach out anyway -- the conversation is usually more interesting than the brief.

A custom build,
on a fixed launch date.

Send a brief, however rough. A first reply lands inside 24 hours on weekdays, usually with two or three questions and a rough shape for the engagement.

Start a Project Booking Q3 2026
© 2026 Carl Mann · All rights reserved · Made by hand on the open web~ delegate the technical ~Set in Instrument Serif, Manrope & JetBrains Mono