Suju Rajan from LinkedIn joined us to talk about how they are operationalizing state-of-the-art AI at LinkedIn. She sheds light on how AI can and is being used in recruiting, and she weaves in some great explanations of how graph-structured data, personalization, and representation learning can be applied to LinkedIn’s candidate search problem. Suju is passionate about helping people deal with machine learning technical debt, and that gives this episode a good dose of practicality.
Matched from the episode's transcript 👇
Suju Rajan: Building off of the standardized version of us looking at the data that we have… Daniel, exactly to your point, we do have a notion of what a member embedding needs to look like. The data for LinkedIn comes from LinkedIn profiles. So in this sense, understanding that “Hey, this particular member has a – in that space, it looks like people who are in this technical industry, in this location, and here are the sorts of jobs that they apply to”, as much as we can represent them in that member embedding space.
Likewise, we have things that are on the job and they’re inside as well; someone asking for a data miner, maybe because they were under a rock, is also still asking for an AI scientist, so how can we bring that job embeddings view into it.
And in some senses, being able to scale this. Because if you think about where – there was an earlier question of “What are the different services in which we have AI?” There is the search aspect of it, and we also organically recommend jobs to people - something that we call instant jobs, where you want to be the first to apply… So members actually sign up for alerts, to say “Hey, the moment you see a job come into our system, then send me an alert right away.” So we need this to be high, high, high precision, as an example.
Then on the flipside, on the recruiter search as an example, it’s all about matching members to get jobs. So we’re trying to build this notion of what we call these two-tower embeddings, which represent the member side and the job side, and bring that together along with more near-line and real-time features, to sort of then personalize it to what that particular member is also looking for, to do ad matching.
Of course, if it’s on the search angle, then you also have the query context that comes into play. On the recommendation side it’s a lot more just based off of your activity signals, and so on and so forth. We try to infer intent, to be able to make those recommendations.
But to your point - yes, on the underlying surface we do have these embeddings, and we are going to be writing a blog post pretty soon about this particular data layer in our talent solution systems.