Adapter AI

Adapter AI turns your existing API documentation into a ready-to-deploy adapter: middleware that speaks the NetApps InsureTech insurer API on one side and your own API on the other. Use it when your API does not already follow the insurer API specification.

Choose a levellink

You choose how much of the platform API your adapter supports when you start, and you can change it later.

To start, open Adapter AI, choose a level and select Start. To change it later, select Change level, choose the new level and select Confirm. Your documents and answers are kept. If analysis has already run, it runs again for the new level, which uses one analysis run.

Follow the five stepslink

The screen shows a status header and five numbered sections. The section you need to work on is marked Current step. Select any section to open or close it.

1. Documentslink

Upload everything that describes your API, then ask for an analysis.

• Accepted files: OpenAPI or Swagger files, Postman collections, Markdown, text or PDF (.json, .yaml, .yml, .md, .txt, .pdf).
• Limits: 25 MB per file and 20 files in total.

Each document shows a status: Upload did not finish, Waiting to be read, Reading, Read or Could not be read. You can remove a document once the AI is not working on your integration.

When everything is uploaded, select Analyse documentation. After you add or replace documents, select Analyse again.

2. Analysislink

Analysis matches your documentation to the platform API and shows your Coverage as a percentage. It sorts each platform item into one of these groups: found as-is in your API, found with a conversion we write for you, found by combining more than one call, not found in your documentation, more than one possible match, or waiting on another answer first.

3. Questionslink

Questions fill the gaps your documentation leaves open. Each question has a priority: Needed to continue, Important or Optional. Open Why we ask and What your documents say for context.

1. Answer the question. You can pick a suggested option, choose "None of these, let me explain" or answer in your own words.
2. Select Save & next, or Skip for now to come back later.
3. Under What we learned, check each statement we took from your answer. Select Confirm if it is right or Not right if it is wrong. When one answer covers several items, Confirm all confirms them together.

4. Mappingslink

Mappings show how each platform API item maps onto your API: which of your calls does it, and which parameters are sent and returned. Each item has a status: Matched, Needs conversion, Combined calls, Missing, Unclear or Waiting. Use Filter by status to narrow the list. An item marked "Out of date. Analyse again." needs a new analysis.

5. Adapterlink

Before we can generate your adapter lists anything still missing, with a button to Answer the question or Complete profile. Then fill in your test system:

1. In Your test environment, enter your test API's Sandbox base URL and select Save test environment. Use the base address only, starting with https://, with no username, password, keys or tokens, and nothing after ? or #.
2. In Test system sign-in, choose how your API authenticates requests: Bearer token, API key in a header, Username and password or OAuth2 client credentials. Enter the values and select Save sign-in details.
3. Optionally, in Extra request headers, add headers your test API needs, such as a tenant or version header, and select Save headers. You can add up to 20.
4. Select Generate adapter (or Generate again for a new build).

Each build shows its start and finish time, its attempts and a status:

Read the status headerlink

After you downloadlink

1. Deploy the adapter at your sandbox API address on the Setup screen.
2. Paste the environment block from the Credentials screen into its settings and restart it.
3. Run Check connection on the Credentials screen.

Owners and admins can run Adapter AI. Notification contacts who chose Adapter build results get emails about your adapter builds.