Job filtering by vector embedding now available + added Apprenticeship job type
jobdataapi.com v4.18 / API version 1.20
Table of contents
vec_embedding filter parameter now available for vector search
In addition to the already existing vec_text filter parameter on the /api/jobs/ endpoint it is now possible to use the same endpoint including all its GET parameters to send a 768 dimensional array of floats as JSON payload via POST request to match for job listings.
This way you're not limited to the vec_text constrains as a GET parameter with only providing text of up to ~1K characters, but can now use your own embeddings or simply those from jobs you already fetched to find semantically similar listings.
With this we now also added a new max_dist GET parameter to be applied optionally to a vec_text or vec_embedding search, setting the max. cosine distance value for the vector similarity search part.
These features are now available on all subscriptions with an API access pro+ or higher plan. See our updated docs for more info.
New Apprenticeship job type added
We saw, for quite a while now, the need to add a job type Apprenticeship to better differentiate certain listings that fall into this category from those that are pure internship roles.
You'll find this popping up on the /api/jobtypes/ endpoint and in relevant job posts from now on (across all API access plans).