Connect a dental practice's on-premises practice management system (PMS) to your AI Employee through Kolla to enable availability checking, appointment booking, and cancellation directly in the practice's schedule. This guide is for Newo.ai partners and technical teams setting up the integration for a practice — budget 30–60 minutes for a first integration.

Supported practice management systems

Kolla connects to on-premises dental PMS software. The connectors Newo.ai uses most reliably are:

Dentrix (Core and Enterprise)
Open Dental
Eaglesoft
Dolphin (orthodontics)

Kolla also lists additional connectors such as WinOMS, OrthoTrac, and DSN, but these are lower-confidence and are not recommended for new deployments without confirming support with Kolla first.

❗❗ IMPORTANT

Kolla reads the PMS database through software installed on the practice's local machine. Cloud-hosted PMS products (for example, the cloud edition of Dentrix) cannot be reached through Kolla. Confirm the practice runs an on-premises installation before scoping the integration.

Prerequisites

Before you start, confirm the following:

A Kolla portal account at app.getkolla.com
An active dental/orthodontic AI Employee

For each practice you integrate, gather:

Practice name and contact email
PMS type and confirmation that it is installed on-premises
The provider(s) and operatory/room the AI Employee should book into by default

Architecture and data flow

Kolla sits between the Newo.ai AI Employee and the on-premises PMS, exposing a single unified API while a connector runs inside the practice network.

Key components

Newo.ai AI Employee : Handles inbound calls, understands patient intent, and manages the scheduling conversation.
Kolla Unified API (cloud) : Provides a standardized REST API across multiple PMS systems.
Kolla connector (on-premises) : Installed on the practice's machine. It reads the local PMS database and connects it to the Unified API.
Dental PMS (on-premises) : The source of truth for schedules and appointments.

Patient identification

The AI Employee identifies callers by phone number (caller ID), matching it against patient records in the PMS. New patients can be booked directly by the agent. For existing patients, most practices prefer changes to be handled by staff, so the agent typically follows the practice's convention and redirects those callers to a manager.

Appointments are tagged so staff can tell how they originated:

(NP) — a new-patient appointment
(EP) — an existing-patient appointment

Supported operations

✅ Check availability
✅ Book appointments (new and existing patients)
✅ Cancel appointments
✅ Patient lookup by phone number
❌ Direct reschedule — modify an appointment by cancelling and rebooking
❌ Multiple locations under a single linked account

All operations are live: appointments the agent creates or cancels appear immediately in the practice's PMS. After a booking, Newo.ai can also email the practice manager on file with the caller's details and the slot booked.

Authentication model

The integration uses three Kolla values:

API token (Bearer token): Authenticates calls to Kolla's Unified API.
Consumer ID : Identifies the practice's linked account in Kolla.
Connector ID : Identifies which PMS connector Kolla should use.

Configure Kolla for the practice

Complete these steps in the Kolla portal for each new practice. You will gather the three required values — API token, Consumer ID, and Connector ID — and the practice will install the on-premises connector.

1. Prepare your Kolla account

Sign in at app.getkolla.com.
Confirm the connector for the practice's PMS is enabled. If it is not listed, contact your Kolla representative to enable it. Invite links can only be created for published connectors.

2. Create a linked account and invite the practice

Go to Linked Accounts and select Create Invite Link.
Choose the connector that matches the practice's PMS.
Enter a unique consumer identifier for the practice. This must be unique across your Kolla account; Kolla does not generate it automatically.
Add the practice (company) name. The contact email field is for your reference only — Kolla does not use it to send the link.
Generate the link and send it to the practice.
The practice administrator opens the link on the machine where the PMS is installed and installs the connector software, which reads the local PMS database and establishes the connection to Kolla.
When the practice completes installation, their record appears as Active in Linked Accounts. Note the Consumer ID from that record.

3. Generate an API token

Go to API Keys and create a new key.
Associate the key with the practice's linked account.
Copy the token and store it securely. You now have all three required values: API token, Consumer ID, and Connector ID.

Important: Remove any leading or trailing spaces when copying credentials. Stray spaces are the most common cause of authentication failures.

Configure your Newo.ai project

With the Kolla values in hand, connect the integration to the customer's AI Employee.

1. Add the Kolla integration module

Open the customer's Builder and go to the Projects page.
Create a new project. Set the IDN and Title to kolla_integration, the Registry to production, and the Module to kolla_integration.
Enable Auto-update and create the project.

2. Load the Kolla attributes

The session is cached when the module is first added, so the Kolla attributes are not visible immediately.

On the new project, open the actions menu (three dots) and select Force Update.
Click Publish.
Go to Attributes and search for kolla to see the Kolla-specific attributes.

3. Enter the required credentials

Fill in the three required attributes (ensure there are no leading or trailing spaces):

kolla_api_token — the API token from Kolla's API Keys page.
kolla_consumer_id — the Consumer ID from the practice's linked account.
kolla_connector_id — the connector identifier you selected (for example, opendental).

Click Save, then Force Update and Publish again. Publishing with valid credentials triggers Kolla to read the practice's system and populate the provider and operatory lists.

Select the default provider and operatory

After publishing with valid credentials, Newo.ai pulls the practice's resources from the PMS.

In the Kolla attributes, confirm that kolla_default_provider (Kolla Provider) is populated with provider names from the PMS, and that kolla_default_operatory (Kolla Operatory) is populated with operatory/room names (if the PMS tracks them).
Working with the practice, select the default provider and operatory the AI Employee should book into.

A provider is a doctor; an operatory is an appointment room. In dental PMS software, providers are usually mapped one-to-one to operatories — for example, a practice might reserve one operatory for new-patient visits. The defaults you select are used whenever the caller does not specify a provider or room.

Tip — provider/operatory names: Practices often refer to providers by first name (for example, "Sarah") while the PMS exposes a resource label (for example, Provider 3). Confirm which resource maps to the named provider before selecting the default.

Attribute reference

Attribute	Purpose	Required
`kolla_api_token`	Bearer token that authenticates requests to Kolla.	Yes
`kolla_consumer_id`	Identifier of the practice's linked account (consumer) in Kolla.	Yes
`kolla_connector_id`	Identifier of the PMS connector to use (for example, `opendental`).	Yes
`kolla_default_provider`	Provider selected from the synced resource list. Used when booking, and as the fallback when dynamic provider resolution is off or unavailable.	Yes
`kolla_default_operatory`	Operatory/room selected from the synced resource list. Used when booking.	If the PMS tracks rooms
`kolla_enable_provider_selection`	When off (default), every appointment uses the default provider. When on, the AI matches the requested service (for example, "cleaning" or "root canal") to a provider by specialty. Keep off until all providers and their specialties are configured.	No
`kolla_appointment_length`	Default appointment length in minutes (default `30`).	No (advanced)
`kolla_appointment_types`	Appointment types supported by the practice, as a comma-separated list of identifiers.	No (advanced)
`kolla_operatories_list`	JSON list of operatory resource names, used to check room availability when calculating free slots. Managed automatically.	No (advanced)
`kolla_providers_catalog`	JSON catalog of providers and their details, used for dynamic provider resolution by treatment type. Managed automatically.	No (advanced)

The advanced attributes are populated and maintained automatically once the integration syncs; you normally do not edit them by hand.

Dynamic provider selection

By default (kolla_enable_provider_selection = off), every appointment is booked with the provider set in kolla_default_provider.

Enable provider selection when a practice has several providers with distinct specialties and you want the AI Employee to route bookings based on the reason for the visit. When enabled, the system uses the providers catalog to match the caller's stated service to the most appropriate provider, falling back to the default provider if it cannot resolve one.

Enable this only after every provider and their specialties are configured correctly in the practice's resources — otherwise bookings may route to the wrong provider.

Scheduling flow

The Kolla scheduling skills are wired into the pre-configured scheduling scenario for dental projects. A typical new-patient booking runs as follows:

Intent recognition — The agent detects that the caller wants to book an appointment and confirms whether they are a new or existing patient.
Information collection — The agent gathers the reason for the visit (appointment type), preferred date and time, and the caller's contact details.
Check availability — The agent gives a brief "let me check availability" acknowledgement while it queries Kolla, which checks the selected provider's schedule (and operatory availability, if rooms are tracked) and returns open slots.
Present slots — The agent offers matching slots; if the requested time is unavailable, it suggests the nearest alternatives.
Confirm and book — The caller selects a slot and confirms. The agent submits the booking to Kolla, which writes the appointment to the PMS, then confirms verbally and can email the practice manager.

For existing patients, the agent follows the practice's convention — typically redirecting booking changes to staff after verifying the caller by phone number.

Test and validate the integration

Run thorough tests before real patients use the integration.

Verify resource sync: Confirm kolla_default_provider and kolla_default_operatory are populated with correct values, and decide whether kolla_enable_provider_selection should be on or off for this practice.
Run an availability check: Call the AI Employee through its normal entrypoint, ask to book an appointment for a known date and time, and verify that the returned slots match the provider's PMS calendar.
Complete a test booking: Select an offered slot and confirm, then check that the appointment appears in the PMS immediately, tagged as a new-patient (NP) appointment.
Test cancellation: Call back and cancel the appointment, then confirm the cancellation is reflected in the PMS.
Review logs: In Newo.ai, review the conversation logs and confirm the scheduling skills fired at the right points.
Repeat multiple scenarios: Complete at least five successful bookings across different dates, providers, and reasons for visit.

Troubleshooting tip: If availability checks consistently return no slots but the calls otherwise run cleanly, the integration is usually working and the provider/operatory selection is wrong. Re-check that the default provider and operatory map to a combination that actually has availability in the PMS.

Once bookings and cancellations consistently appear in the PMS as expected, treat the integration as production-ready for that practice.