Kirnova developer documentation
Kirnova provides TIGER, a framework that reads how a person works across five dimensions: Trust, Influence, Grit, Execution and Resilience. You can deliver Kirnova assessments inside your own product and receive each result on your servers.
This portal contains the written guides, an interactive API reference and the OpenAPI specification.
What you can build#
- An embedded assessment. Place a Kirnova frame in your product. Kirnova renders the questions, scores the answers and sends you the result.
- A fully custom sitting. Drive an assessment from your own servers and render every screen yourself.
- A results pipeline. Receive signed completion messages, fetch a single result, list results for a period, and request a prepared document where one is available.
Which assessments you can run, and on which surface, is covered in Capabilities and limits.
Get access#
- Contact team@kirnova.com with the assessments you want to run, the surface you plan to use (embedded frame or machine-driven API) and what you will use the results for.
- Kirnova creates your account and enables the assessments you are entitled to. An account with no assessments enabled cannot start a sitting.
- Kirnova issues a staging key. Once your account exists, you can create and revoke your own keys.
- Register the address that receives results and, for the embedded frame, the sites you will embed on.
Test your integration#
Build and test on Kirnova staging at https://stage.kirnova.com with a staging key. Staging has its own database, and its data is test data. Keys do not work across environments: a staging key is refused on production, and a production key is refused on staging. Staging keys can look like production keys, so use the address to tell the two apart.
Production endpoints are https://api.kirnova.com for the API and https://embed.kirnova.com for the frame. The integration playbook contains a test plan to run on staging before you go live.
Guides#
| Guide | What it covers |
|---|---|
| API guide | Every endpoint, field, event, error and limit |
| Integration playbook | The order to build in, with sequences, a worked example, a test plan, troubleshooting and go-live steps |
| What Kirnova delivers | The assessments, the five dimensions, and who may start each assessment |
| The respondent journey | What the person sees, every way a sitting can end, and what to tell a person who contacts you |
| Reading a result | What a result contains, what it means, and how to display it accurately |
| Capabilities and limits | What is available, on which surface, and what is not |
| Use cases | Common scenarios, with what you do, what you receive and where the limits are |
Principles that apply to every integration#
- Show a score with its range. Every reading carries a range. A number without its range implies more precision than the assessment provides.
- Treat overlapping ranges as level. Readings are not percentiles. If two people's ranges overlap, the assessment cannot tell them apart.
- TIGER informs; people decide. A result can inform an employment decision. The product has no pass mark and no automatic rejection, and your integration must not add either.
- Identify every respondent. There is no anonymous sitting. Every respondent is identified before the sitting starts.
Support#
For a new assessment, a higher limit, a missing result or anything that differs from these guides, write to team@kirnova.com. For a problem with a sitting, include the session identifier and the approximate time it happened.