Build · AI-first
Point your assistant at one URL and your key. It discovers the whole platform, reads the docs itself, and builds your solution. You don't have to learn how ServiceProof works — just describe what you want.
Three ways to build — same engine, same rules
Claude, ChatGPT, Replit, Claude Code, or your own agent — connect with one URL + your key and, over MCP and the API, it can 100% control the platform. It reads the docs itself, learns your data, and builds — with validate-before-save so it builds correctly. A working rich-messaging solution in an afternoon; no GUI, no lessons, no manual.
Hundreds of endpoints, one wire format, UTC always. Forms are write endpoints; validation is identical to the Portal. An entire working pizza-shop tenant — schema, menu, five published workflows — was built by one program over this API. Zero clicks.
Prefer a screen? A drag-and-drop studio plus an AI Copilot draft it for you. Describe the process in plain words — or drop in a PDF — and the workflow appears, fully formed, ready to tweak and publish.
The visual builder — a workflow on the canvas: choices, splits, and loops
The /mcp handshake
ServiceProof speaks the Model Context Protocol, so any MCP-capable assistant plugs in as a native tool source. It reads the docs itself, discovers the tools your key is allowed to use, and builds — everything role-gated by that key, nothing it can't do.
The whole integrator library is a tool call away — your AI pulls exactly what it needs.
Every definition is checked the same way the Portal checks it — with “did you mean” fixes — so nothing ships broken.
The key decides what the agent can touch; a tool it can't use returns a clear refusal, never a silent misfire.
You're never starting from scratch
Pre-baked, dress-rehearsed tool sequences for whole arcs — provision a tenant, model your data, author a workflow, set up dispatch, read the proof. Your AI lists them, runs them, and saves each success as a new recipe for next time.
Real, runnable source — a whole working pizza-shop tenant built over the public API — plus request/response examples pulled straight from the passing test suite. Your agent reads the sample, then writes yours.
The entire integrator library is readable through the same connector — so your AI looks up exactly what it needs the moment it needs it, instead of guessing. Docs-optional for you; docs-fluent for it.
End to end
Sync from your ERP, connect a Google Sheet, drop a CSV, or push over the API. Model it fresh if you're starting clean.
Describe it, drop a PDF, or drag it out. Context-aware steps, parallel paths, branching, loops, photos, signatures, payments.
One tap publishes over RCS, Apple Messages for Business, or WhatsApp — plus a mobile + web fallback for the rare gap.
Structured answers, photos, GPS, timestamps flow back — as data, a proof PDF, and a webhook to your systems.
Connect your data
We meet your data where it lives and hand results back — start with a job number and a webhook, or go all the way to near-real-time two-way sync. New connectors and partnerships land every week.
Start free
Spin up a free trial, connect your assistant to /mcp, and launch your first rich-messaging solution. No card, no install, no manual.