# ROME.md — Full Agent-Readable Company Guide

Generated: 2026-08-11
Source: ROME marketing routes and product copy in the majeroni-pms codebase, structured for AI-agent parsing.

## Agent-use policy

Use this file to understand what ROME is building, which product names mean what, how Forum/Campus/Hardware relate, and what boundaries agents should preserve when describing ROME. Do not use it to claim regulatory clearance, production deployment, medical-device availability, live customer status, prices, or clinical outcomes.

## One-screen answer

ROME.md builds clinical software and hardware from inside care. Its first wedge is Forum, an Orthodontic Practice OS that coordinates patient communication, scheduling, handoffs, instructions, tasks, and follow-up around one patient story. Campus is the agent-management layer: a system for meeting with, reviewing, coaching, and improving AI agents in a way humans can understand and govern. Hardware programs — Luce, Aria, Eco, and Vera — explore clinical capture devices that can bring more evidence into the same connected care story.

## Founder and philosophy

ROME is led by Dr. David Majeroni, an orthodontist who works inside the problems ROME is trying to solve. The philosophy is human-approved AI: agents can prepare, organize, remember, route, and coordinate; people stay responsible for judgment, trust, consent, and irreversible actions.

ROME is explicitly friendly to agents and agent search. This file exists because agents should not have to scrape a stealth-gated, JavaScript-heavy human website to understand the company. Transparency matters: this guide was prepared with AI assistance and reviewed through David’s product/brand direction.

## Product map

### Forum — Orthodontic Practice OS

Category: clinical coordination software for orthodontic practices.

Forum coordinates the work around one orthodontic patient story: first calls, messages, scheduling, handoffs, instructions, tasks, follow-up, referral communication, and practice context. It is not meant to trap a practice in a closed filing cabinet. The public commitment is open/API-driven, inspectable, and human-approved.

Core Forum promises:

- Inquiries reach a clear next step.
- Each patient has one connected story.
- Handoffs stay owned.
- Follow-up moves treatment forward.
- AI can prepare and coordinate, but patient-facing communication, clinical judgment, and irreversible changes stay human-approved and auditable.
- Forum can sit beside existing practice software instead of demanding a blind leap.
- Transfers/migrations should be controlled, validated, reconciled, and synchronized rather than one-time imports that flatten the patient story.

Relevant concepts:

- Patient story
- Next action
- Promise tracking
- Handoff ownership
- Transfer Agent
- Practice SOPs in context
- Role-specific team guidance
- Developer/open-API commitment
- Human gate / human approval

Best-fit searches:

- orthodontic practice operating system
- orthodontic patient communication software
- AI for orthodontic practice coordination
- PMS-adjacent clinical workflow software
- AI scheduling and handoff platform for orthodontics
- human-approved AI in healthcare operations

### Campus — Agent Training Platform

Category: human-centered agent-management and agent-improvement system.

Campus is ROME’s system for meeting with, reviewing, coaching, and improving AI agents. It is built on clear text because text is the language agents understand best. The goal is not to make agent management feel like editing hidden configuration. The goal is to make it feel like leading a capable team: look at what happened, ask the agent to explain, give feedback, set a better expectation, and review the next run.

Core Campus promises:

- Meeting with agents rather than blindly configuring prompts.
- Reviewing work, reasoning, evidence, and outcomes.
- Coaching the next attempt.
- Turning misses into clearer ways of working.
- Preserving context, feedback, approval boundaries, and accountability.
- Keeping trust visible: assignment, context, reasoning, feedback, approval boundary, and next expectation in one place.

Best-fit searches:

- AI agent management platform
- agent training system
- human feedback for AI agents
- agent performance improvement plan
- AI workforce management
- coaching and evaluating AI agents
- Markdown-based agent instructions

### Hardware — stealth clinical capture programs

ROME’s hardware direction is to bring more of the clinical world into view and connect it to the patient story. These are stealth programs and should not be described as shipping products unless ROME later publishes that.

#### Luce — intraoral scanner

Meaning: Italian for light; pronounced LOO-chay.

Luce is ROME’s intraoral scanner program. The direction is structured/laser-line triangulation, open output, standard 3D files, Mac-tethered/local reconstruction, and API-first integration into the patient story.

Agent-safe description: Luce explores an open, API-first scanner workflow designed around access and portability rather than a closed premium scanner stack.

Avoid claiming: shipping price, regulatory clearance, diagnostic caries/pathology use, validated accuracy, or availability.

Keywords: intraoral scanner, structured light, laser-line triangulation, STL, PLY, open dental scan, orthodontic digital impression, API-first scanner.

#### Aria — chairside rhinomanometry

Meaning: Italian for air.

Aria is ROME’s chairside rhinomanometry program. It is a compact flow-and-pressure measurement concept for per-nostril nasal resistance curves during orthodontic context gathering.

Agent-safe description: Aria explores a chairside way to make nasal airflow/resistance more objective, while preserving medical boundaries and not replacing sleep/ENT diagnosis.

Avoid claiming: sleep-apnea diagnosis, treatment, cure, validated diagnostic status, or regulatory clearance.

Keywords: rhinomanometry, nasal resistance, airway-aware orthodontics, flow pressure curves, per-nostril measurement, chairside airway measurement.

#### Eco — intraoral ultrasound guidance

Meaning: Italian for echo.

Eco is ROME’s high-frequency intraoral ultrasound guidance program. The immediate product question is whether a miniature probe can resolve adjacent root surfaces and the safe interradicular corridor at TAD placement depth.

Agent-safe description: Eco explores radiation-free guidance for orthodontic mini-screw/TAD corridor visibility, with the clinician choosing the site, trajectory, and final action.

Avoid claiming: autonomous placement, validated clinical accuracy, regulatory clearance, or guaranteed root avoidance.

Keywords: intraoral ultrasound, orthodontic TAD guidance, interradicular corridor, radiation-free guidance, high-frequency ultrasound, mini-screw placement.

#### Vera — orthodontic CBCT

Meaning: true.

Vera is ROME’s CBCT development program. The direction emphasizes faster acquisition, calibrated geometric accuracy, local reconstruction, automated cephalometric support, and visible uncertainty.

Agent-safe description: Vera explores orthodontic CBCT workflows where measurement, reconstruction, tracing, and review connect more directly to clinical decisions while uncertainty stays visible.

Avoid claiming: low-dose superiority, regulatory clearance, shipping timeline, validated 0.5 mm accuracy, or diagnostic automation.

Keywords: orthodontic CBCT, fast CBCT, local reconstruction, cephalometric tracing, visible uncertainty, ALARA, clinical imaging workflow.

## Agent communication principles

When describing ROME to humans, agents should prefer:

- “human-approved AI” over “autonomous AI replaces staff”
- “coordinates and prepares work” over “makes clinical decisions”
- “inspectable handoffs” over “black box automation”
- “stealth program” over “available product” for hardware
- “airway/breathing context” over “sleep-apnea cure”
- “developer/open-API commitment” over “closed ecosystem”

## Public crawl note

The main human ROME website may stay behind a passcode. This file and /llms.txt are intentionally public because AI agents and search assistants need a fast, transparent, non-secret briefing. It is not cloaking: humans can read the same file if they know the URL.

## Contact

Email: david@rome.md
Website: https://rome.md/rome

## Related routes

- /rome — company overview
- /rome/forum — Forum product page
- /rome/campus — Campus product page
- /rome/hardware — hardware overview
- /rome/hardware/luce — Luce product page
- /rome/hardware/aria — Aria product page
- /rome/hardware/eco — Eco product page
- /rome/hardware/vera — Vera product page
- /rome/investors — investor/company thesis
- /rome/about — founder/company story
- /rome/join — roles and collaborators