Skip to main content
You run a clinic on Tebra, and you’ve built a Medical Agent in A2V2.ai to help patients. But your patients and their history live in Tebra, not in A2V2.ai. The Tebra import brings them across: it reads the clinical export files from Tebra and turns each patient into a CRM contact — with their notes, medications, vitals, and labs attached — so your agent and your team work from the same records. Open Integrations → Tebra from the dashboard sidebar, then Open import wizard. The wizard is titled Import from Tebra, with the subtitle “Step-by-step guided import of CCDA dumps from Tebra.”
Availability is gated, and this is healthcare-specific. Tebra import is a selectively rolled-out, healthcare-vertical feature. If you don’t see Integrations in your sidebar, it isn’t enabled for your organization — contact A2V2.ai support. Because it moves real patient data, it’s restricted to the organization Owner.

​
Prerequisites

  • You’re the organization Owner (the screen is Owner-only).
  • Integrations is enabled for your organization.
  • A target agent to attach the imported patients to — typically a Medical Agent (see Create an agent).
  • Your Tebra Organisation ID (and optionally its name).
  • The CCDA export files from Tebra — the standard clinical-document export that carries each patient’s record.
What’s a “dump”? The wizard calls one batch of Tebra export files a dump and gives it a Dump ID. Think of it as one import run — everything you bring across in a single pass. You can have several over time, and the Recent imports list lets you resume an unfinished one or review a past one.

​
How the import works

The wizard walks through six labeled steps. It deliberately splits the work into two stages: Stage 1 reads and checks the data (nothing is created yet), and Stage 2 writes it into A2V2.ai. That gap is your chance to review before anything lands in your CRM.
1

Configure

Turn on Enable Tebra integration, then enter your Tebra Organisation ID (and an optional Tebra Organisation Name). Choose a Default Provider User — the A2V2.ai user that patient records fall back to when their Tebra provider isn’t mapped to a specific person. Select Save settings, then Continue.
2

Create dump

Pick the Target chatbot the patients will belong to, then Create dump. The wizard allocates a new Dump ID for this run and prepares it to receive your Tebra export files.
3

Parse (Stage 1)

Start Stage 1. A2V2.ai reads every patient file, validates it, and extracts demographics, problems, allergies, medications, vitals, labs, encounters, and clinical notes into a structured staging area. No A2V2 records are created in this stage — it’s read-and-prepare only. You’ll see a running “Patients parsed” count; if any patient fails to parse, you can retry just those before continuing.
4

Verify & review

The wizard runs a preflight checklist and shows a dump preview so you know exactly what’s about to be imported — total Patients, any with Missing contact info (these get skipped, because A2V2.ai needs at least an email or phone to create a contact), Repeat patients already in your CRM (these are refreshed, not duplicated), and the Encounter date range. Anything that would block the import is flagged here, and you map any unmapped Tebra providers (see below) before moving on.
5

Map (Stage 2)

Start Stage 2. Now A2V2.ai writes the parsed data into your workspace: one CRM contact per patient, with clinical notes, prescriptions, vitals and labs, and any attached patient files. Re-running is safe — repeat patients update in place, so you won’t get duplicates.
6

Done

You get an import receipt summarizing what landed: patients imported plus counts of clinical notes, active prescriptions, vital/lab results, patient files, form submissions, and tasks — along with any patients that were skipped or failed. From here you can View patient records, Print / save as PDF for your records, or Start new import.

​
Provider mappings

Tebra identifies each provider by an NPI (National Provider Identifier). So that imported patients are attributed to the right person in A2V2.ai, the wizard lets you map each Tebra NPI to an A2V2 user during the Verify & review step. You can also manage these later from Provider mappings on the Tebra screen.
If a provider’s NPI has no explicit mapping and you haven’t set a Default Provider User, those patients’ records will fail to import. Either map each NPI or set a fallback user in Configure before running Stage 2.

​
Managing provider mappings

Mappings are first created during the wizard’s Verify & review step, but you don’t have to re-run an import to change them. Open Manage provider mappings from the Tebra screen for a standing view of every provider A2V2.ai has seen — titled Tebra provider mappings, with the subtitle “Edit or remove the A2V2 user assigned to each Tebra provider NPI.” Like the rest of Tebra, this screen is Owner-only. At the top, four tiles summarize the current picture — Total providers, Mapped to user, Using fallback, and Skipped — so you can tell at a glance how many providers still route through the default user. Below them, a search box lets you filter by NPI, name, or A2V2 user, and the table lists each Tebra NPI, its assigned A2V2 user (or Not mapped), and its Strategy.
Removing a mapping doesn’t touch imported records. When you remove a mapping, patients you’ve already imported stay exactly as they are — only the NPI’s future routing changes. The provider simply shows up again as unmapped the next time you run an import, so you can re-assign it then.

​
Recent imports

The Tebra screen keeps a Recent imports list so you can pick up where you left off or look back. Each row shows the Dump ID, Chatbot, Status, Progress, and when it was Started. From a row you can Resume an in-progress import, View a finished one, or Discard it. Discarding cancels a dump and drops its staged data, but records already imported into your CRM are not removed by discarding.

​
Patient records for an import

An import finished, but one patient you expected isn’t in your CRM — did they not come across, or did something go wrong for just them? The Patient records screen answers that. Open it with View patient records from the import receipt (or from a finished dump) to see one row per patient in that dump, with exactly where each one stands. Like the wizard, this view is Owner-only, because it shows per-patient detail. The screen is headed Patient records, described as “All patients in this dump with their per-section status.” Each patient shows two independent status columns — Stage 1 (did we read the file?) and Stage 2 (did we write the records?) — so you can see at a glance whether a patient stopped at parsing or at mapping. A Partial Stage 2 means some of that patient’s clinical sections imported while others failed — their contact exists, but part of the record didn’t map. To zero in on problems, use the two filters at the top: filter by Stage 1 status, Stage 2 status, or both (each has an All option), for example Stage 2 = Failed to list only the patients that didn’t map. Select any row to open its detail drawer, which breaks the patient down into Patient detail, Parsed CCDA (the structured data read in Stage 1), Mapped records (links to what was created in your CRM), and Errors (why a section failed, if it did). When a patient failed or came in Partial, the Errors section is where you find out why.
Staged clinical data is cleared after a successful import — by design. Once a patient imports successfully, A2V2.ai automatically redacts the raw clinical export it staged, and the Parsed CCDA section of the drawer shows “Parsed CCDA redacted.” This is expected, not an error: the imported records live on in your CRM, and the drawer still links to the Mapped records so you can open them there. Only the temporary staged copy is removed.

​
What ends up in your CRM

After a successful import, each patient is a CRM contact. Depending on what their Tebra record contains, the import can also create or refresh:
  • Clinical notes — allergies, problems, encounters, procedures, and history.
  • Active prescriptions — from the patient’s medications.
  • Vital signs and lab results — stored with their value history.
  • Patient files — lab PDFs, intake forms, and images.
  • Form submissions and plan-of-care / referral tasks.

​
Tips

Run Stage 1, then review before Stage 2. The two-stage design exists so you can catch problems — skipped patients, unmapped providers, unexpected repeat counts — while nothing has been written yet. Treat Verify & review as a real checkpoint, not a formality.
Fix missing contact info in Tebra, not after. Patients with neither an email nor a phone are skipped, because A2V2.ai needs at least one to create a contact. If the preview flags skips you care about, add the details in Tebra, re-export, and run a new dump.
Re-importing is safe. Bringing the same patient across again updates their existing record in place rather than creating a duplicate, and their notes, prescriptions, and clinical data carry over.

​
Troubleshooting

Integrations is a selectively enabled, healthcare-specific capability, and it’s Owner-only. If you’re the Owner and still don’t see it, it isn’t enabled for your organization — contact A2V2.ai support.
Only the organization Owner can run Tebra imports. Ask your Owner to run it, or review your role.
Patients are skipped when their Tebra record has neither an email nor a phone — A2V2.ai needs at least one to create a contact. Add the missing details in Tebra, re-export, and run a new import.
The most common cause is an unmapped provider with no fallback user. Map each Tebra NPI to an A2V2.ai user, or set a Default Provider User in the Configure step, then retry the failed patients. The wizard lets you retry just the affected patients rather than re-running everything.
No. Stage 2 is idempotent — a patient who already exists in your CRM is updated in place, and the preview’s Repeat patients count tells you how many will be refreshed before you start.
Open that dump’s Patient records (View patient records from the receipt or a finished import), then filter Stage 2 status to Failed or Partial. Select the patient’s row and read the Errors section of the detail drawer to see which section didn’t map and why.
That’s expected after a successful import. A2V2.ai automatically clears the raw clinical export it staged for the patient once their records are created. The imported data lives in your CRM — open the Mapped records links in the same drawer to inspect it there.
Open Manage provider mappings from the Tebra screen, find the NPI, and set its Strategy to Exact with the correct A2V2 user — or set a Default Provider User in the Configure step so unmapped providers fall back to a known owner.

Integrations overview

What integrations are and how to reach them.

CRM contacts

Where imported patients live — view, edit, and work with them.

Create an agent

Set up the Medical Agent your patients attach to.

Members & Roles

Why Tebra import is Owner-only, and how roles work.