Import or start guided
Choose either guided path: import prior schedules or exported CSV files and review the records AneSched infers, or begin with blank databases while the AI wizard helps the administrator complete each one in turn.
Built for anesthesia scheduling
Talk through the schedule. Approve the rules.
AneSched turns the way your group actually works into reviewable records, validated rules, and a monthly schedule—without asking an administrator to become a programmer.
Current MVP: physician-only or CRNA-only anesthesia groups. Mixed care-team models involving physicians with CRNAs, CAAs, or residents are planned.
My interpretationUpdate Dr. Maya Rivers
94% confidenceCardiac capability Qualified
Institution eligibility Pacific Memorial
OB callSoft preference: avoid
This is a nudge, not a hard prohibition. The validator may still use Dr. Rivers when necessary.
Edit interpretation
Nothing changes silentlyApprove the interpretation, then Save the canonical record.
Conversation firstExplain your group in plain language
Human approvedEvery canonical change is confirmed
Deterministically checkedHard constraints remain predictable
Locally controlledYour canonical schedule database stays on your computer
From old schedules to a new month
The AI learns the shape of your group with you.
Begin with historical schedules or a guided blank database. AneSched proposes what it sees; the administrator remains the source of truth.
Resolve uncertainty
Assumptions below 80% confidence receive a question mark. The AI asks focused questions at the record level instead of turning setup into a field-by-field interrogation.
72% Needs administrator review
Approve each record
Review Dr. Rivers, Pacific Memorial, or 1st Call Night as a complete record. Correct the AI in plain language, approve its interpretation, and press Save to make the canonical change.
Dr. Maya Rivers Approved
Build, inspect, refine
The scheduling engine applies approved hard constraints, uses soft preferences as nudges, and presents a proposed month for administrator review. Point out a problem; AneSched translates the correction into a governed change.
AI is the collaborator
The rule engine is a guardrail—not the personality of the product.
AneSched is designed so administrators can describe exceptions, preferences, and local scheduling culture naturally. The AI maps those instructions to capabilities the scheduler already supports and tells you when a request would require a governed product upgrade.
- Hard requirements stay enforceable and testable.
- Soft preferences influence fairness without becoming brittle rules.
- New group-specific behavior is proposed, tested, versioned, and reversible.
The capability boundary
Supported today
“Avoid OB call for Dr. Rivers when reasonable.”
+
Needs a governed upgrade
“Create a new three-team supervision model.”
The restricted Codex agent explains which side of the line a request falls on before anything changes.
In-product beta support
Report the problem where it happens. Keep the fix governed.
An issue report inside the MVP leads to one clear next step: guidance for a capability the installed product already supports, a focused clarification, or a governed product-update candidate for review.
Report inside AneSched
Describe what happened and the result you expected. The administrator-controlled workspace records the report.
Receive the right path
AneSched returns current-product guidance, asks a focused question, or identifies a product-update candidate.
Approve a tested release
When code must change, development happens outside the installed scheduler. The change is tested, versioned, and presented for approval before release.
Fast improvements. No silent changes.The running scheduler does not rewrite itself, transmit scheduling data automatically, or install an update on its own.
Designed for responsible control
Your schedule database is not stored on this website.
The website explains and distributes the product. The canonical database and validators remain in the administrator-controlled installation. When an administrator deliberately uses an AI feature, AneSched may send the bounded context required for that task to OpenAI through the connected ChatGPT account.
On Cloudflare
AneSched.com
Product information, recorded walkthrough, verified release downloads.
No schedule databaseAdministrator controlled
Local AneSched workspace
Canonical database, deterministic validators, and a restricted native Codex agent connected through ChatGPT OAuth.
Local canonical databaseExplicit approval
AI-proposed changes do not become canonical until the administrator confirms and saves them.
Restricted AI operator
The Codex agent is bound to AneSched’s approved roles and cannot rewrite the installed program.
Auditable changes
Approved rules and record edits are versioned so administrators can understand what changed.
Early Access Beta · MVP validation in progress
One download contains the complete Windows 11 x64 beta.
Authorized evaluators download one Windows 11 x64 ZIP containing AneSched, its required local runtimes, the restricted native Codex agent, and the guided setup. Nothing else needs to be downloaded separately. Extract the folder, then double-click START-ANESCHED.cmd.
A ChatGPT account with Codex access is required. OAuth provides the secure sign-in; the connected ChatGPT plan and the amount of AI use determine available capacity. The Windows computer runs the local AneSched service and must be awake while AneSched is in use.
First-release scopePhysician-only or CRNA-only anesthesia groups
Two guided onboarding pathsReview records inferred from old schedules, or build the blank databases with the AI wizard
Codex usage visibilityPlan, percentage used, percentage remaining, and next reset date and time in AneSched
Deterministic quality gateBackend, frontend, installer, and scheduling validation
Verified internal packageOne versioned ZIP with a published SHA-256; signing and controlled upgrades remain future release gates
Protected beta download The current unsigned development ZIP remains limited to authorized evaluation and is not approved for clinical or production use.
Open the one-download page Straight answers
What an anesthesia administrator should know.
Is the website a live scheduling demo?
No. The website shows a recorded, fictional walkthrough. The real experience comes from installing the MVP and signing the restricted Codex agent in through ChatGPT OAuth.
Which ChatGPT plan do I need?
You need a ChatGPT account with Codex access. The right plan depends on how often and how heavily the group uses AI. AneSched shows the administrator the percentage used, percentage remaining, and the next reset time so capacity can be reviewed before changing plans.
Can the AI change the database?
Not directly. It returns its interpretation in plain language, the administrator approves or corrects it, AneSched validates it, and an explicit Save makes the canonical change.
Where is scheduling data kept?
The canonical schedule database remains on the administrator’s computer. An intentional AI request may send only its bounded task context to OpenAI through the connected ChatGPT account. This beta does not yet claim qualified portal backup, mobile continuity, or replacement-computer recovery.
What happens when the import is uncertain?
A value below 80% confidence is marked for attention. The AI asks a focused question, then the administrator approves the completed staff, institution, or position record.
Does it support anesthesia care-team supervision?
Not yet. The current MVP can schedule physician-only or CRNA-only groups. Mixed care-team models involving medical direction or supervision of CRNAs, CAAs, or residents are planned.