Services · V3 · Conversation Systems

Conversation Systems

Your people need to rehearse the exact conversations you can't send to someone else's server.

Every funded player in conversation practice is cloud SaaS priced per seat. But the conversations people most need to rehearse — a disclosure, a discipline, a difficult diagnosis — are exactly the ones you can’t send to a server. Conversation Simulator is local by architecture: there is no server, and you can verify it with one command. That’s a structural advantage in any regulated or privacy-bound setting, and the YAML scenario format makes content a product line, not a feature.

The offers

Scenario Pack Commission

$6,500 fixed2–3 weeks
  • 5–8 professionally authored scenarios for your specific practice need
  • Grading rubrics and debrief logic, not just dialogue
  • Safety-reviewed against the two immutable rules the engine ships with
  • Delivered as YAML you own, running on hardware you control

Sized to slip under a $10k discretionary threshold.

Ask about the Scenario Pack Commission ›

Private Practice Pilot

$12,0006 weeks
  • A commissioned scenario pack, branded and deployed for your team
  • A facilitator guide so a non-engineer can run the sessions
  • Pre/post measurement design for one cohort
  • Priced deliberately under a year of a 50-seat cloud contract — with zero data egress

One cohort of up to 50 practitioners.

Ask about the Private Practice Pilot ›

Conversation Simulator

$9.99 on SteamFree from source
  • The shipped consumer product: practice interviews, negotiations, language, and hard conversations
  • Apache-2.0 engine you can audit
  • An offline smoke test that proves the no-egress claim
  • The same architecture your institutional pack runs on

Runs 100% on your computer — no account, no cloud, no telemetry.

Get Conversation Simulator ›

Proof, not adjectives

See everything under this vector ›

For the compliance file

The privacy architecture is the compliance answer: local models, an on-device SQLite database, and a verifiable offline receipt. We wrote it up as a one-pager your privacy office can read on its own — the local-first stance mapped to HIPAA and FERPA postures in plain language (descriptive, not legal advice).

Read the regulated-sector one-pager — or hand your procurement team the PDF:

Download the regulated-sector one-pager (PDF)PDF · 2 pages · print-ready · HIPAA / FERPA postures in plain language

How it works

A Scenario Pack Commission runs on measurable exit criteria — the debrief rubric is the graded artifact, not model fluency:

  1. Scope the need. Sector, sensitivity level, cohort size, deployment constraints — we define what “good practice” looks like for you. Exit: a one-page scenario brief you sign off.
  2. Author and review. Scenarios, rubrics, and debrief logic written and safety-reviewed. Exit: every scenario passes workbench validation and the content-review policy.
  3. Deploy on your hardware. The pack runs where your data already lives — no server, no egress. Exit: a facilitator runs a full session on your own machines.
  4. Measure. Pre/post design so the cohort’s improvement is a number, not a vibe. Exit: a measurement plan your team can report on.

Fit / not-fit

This is for you if

  • You train people in healthcare, legal, social work, education, or another privacy-bound field
  • “The data never leaves the building” is a compliance requirement, not a nice-to-have
  • You want scenarios your institution owns and can edit, not a per-seat subscription
  • You’d rather grade the debrief than trust a fluent-sounding model

We're probably not your studio if

  • You want a cloud dashboard with per-seat analytics and SSO across 10,000 users
  • You need the model to sound like a specific real person
  • You’re buying a one-off demo with no intention of running real cohorts
  • You want the lowest per-seat price and don’t value the local-first architecture

Scenario-need intake

Four questions so we scope the right thing — the consumer app, a measured pilot, or a commissioned scenario pack your institution owns and runs on its own hardware.

Prefer to talk first? Ask for a 20-minute fit call in the form and we’ll send a time.

This is why local-first matters — the more sensitive, the more the no-egress architecture earns its keep.

The practice need (optional)
Where to reach you
No form? Send it as an email instead.

Copy the template, fill in the blanks, and send it to hi@outrightmental.com. One person reads every note and replies within two business days.

Open this in your email app ›

Want the consumer relationship version? → readtheroom.fyi/

Questions people ask

How do you price this against cloud competitors?
A 50-seat cloud conversation-practice contract runs roughly $18k–$30k a year at market per-seat rates. A Private Practice Pilot is $12k, once, for a commissioned pack you own and run on your own hardware — and the commission is sized to fit under most discretionary-spend thresholds so you don’t wait on procurement.
Who owns the scenario packs?
You do. Commissioned packs are delivered as YAML you own outright, running on hardware you control. The engine is Apache-2.0; the packs are yours. There’s no lock-in, because there’s no server to be locked into.
How is safety handled in commissioned content?
The engine ships with two immutable safety rules and a validation workbench. Commissioned work adds a published content-review policy on top, and every scenario is safety-reviewed before delivery. Sensitive material is exactly what this architecture is built to handle responsibly.
What happens after the pilot?
You keep the pack, the facilitator guide, and the measurement design — the pilot stands on its own. If the numbers justify it, year two is a site-license conversation informed by your cohort’s actual results, not a projection.