View All Case Studies

CUSTOMER SUCCESS STORY

How AMN Healthcare Scaled Personalized Job Recommendations with an Advanced Recommendation Recipe

AMN Healthcare built an advanced recommendation recipe with Blueshift to deliver personalized job recommendations that stay relevant as clinician preferences, location constraints, and job inventory change.

Cover for AMN Healthcare case study featuring the AMN Healthcare logo with a blurred image of a healthcare team in the background
10.7M+
Total Sends
1.5M+
Total Clicks
~20%
Avg Open Rate
15%
Peak Click Rate

The Challenge

Clinicians come to AMN Healthcare with a clear expectation. They don’t want to sift through thousands of job postings. They want to feel understood. They want roles that match their training, their specialty, and the places they’re willing to work. With a talent marketplace as large and diverse as AMN’s, that level of precision isn’t a luxury. It’s table stakes.

AMN knew that delivering true 1:1 job matching would take more than a basic recommendation engine. They needed a flexible framework that could understand clinicians as individuals, scale across millions of job interactions, and update itself as needs and behaviors changed.

That’s what set the stage for their partnership with Blueshift and the creation of a new, advanced recommendation recipe for personalized job recommendations.

The Outcome

AMN built a recommendation pipeline designed to turn clinician preferences and behavior into ranked job matches that feel practical, not generic.

A perfect role loses its value if it’s nowhere near where a clinician can or wants to work. So the recipe brings location intelligence into the equation and applies relevance boosts, business rules, and intelligent fallback logic to keep recommendations useful even when the personalized pool is narrow.

How the Advanced Recommendation Recipe Works

Here’s how a robust recommendation pipeline works: from understanding clinician preferences to delivering ranked job matches.

blank

Step 1: Identify Clinician Job Preferences
The journey began with a simple but essential question: Who is this clinician, and what are they actually looking for? This is the foundation of the entire personalization experience. This initial stage captures the essential dimensions of a user’s career aspirations. The recipe evaluates what clinical skills the user specializes in, their professional category, and any sub-specialization that helps to identify the most relevant roles.

Step 2: Match Category Jobs to User Preferences
Once the user profile was clear, the next challenge was scale. AMN has thousands of open roles, and the system needed to sift through all of them to find the ones that made sense for each clinician. This categorical matching formed the first real layer of personalization. It filtered the entire job universe down to a curated set of opportunities aligned to each clinician’s background.

But “clinically relevant” wasn’t enough. The recommendations had to feel practical.

Step 3: Putting Location Intelligence to Work
A perfect role loses its value if it’s nowhere near where a clinician can or wants to work. This stage brings geographical intelligence into the equation. The workflow applies two types of location signals: first, the location the clinician has explicitly added to their profile, and second, the system’s understanding of where the user tends to interact, based on past engagement (if available). This geo layer ensures recommendations are not just professionally relevant but also where the clinician is both able and willing to work.

Step 4: Enhance Results with Relevance Boosts
At this stage, AMN had relevance. Now they needed resonance. This stage applies intelligent ranking factors to surface the most compelling opportunities. Recency ensures users see newly added opportunities. Hot Jobs highlights high-demand or priority jobs. Filtering rules apply business logic and quality controls. This enhancement layer transforms a simple list of matches into a curated selection of high-quality opportunities, aligned with AMN’s priorities.

Step 5: Intelligent Fallback Logic
If the personalized pool is too narrow (e.g., few matching roles are available), the system automatically expands the scope to ensure meaningful recommendations are always shown. This step guarantees consistency: every clinician receives a rich set of job options, even with limited profile data or inventory constraints.

Delivering the Final Experience

The output of this pipeline is a personalized experience that feels curated, prioritized, and intelligently diversified. Tailored to their background, location, inferred interests, and automatically refreshes as new jobs appear or user behavior changes. This systematic approach means users spend less time wading through irrelevant postings and more time engaging with opportunities that could genuinely advance their careers. For job platforms, it means higher engagement and better matches. 

A Blueprint for Scalable Personalization

AMN Healthcare’s personalization journey is a perfect example of how personalization, when built on a flexible, adaptive foundation, simplifies everything for marketers, clinicians, and the business. They’ve created a system that’s both powerful and easy to maintain, delivering deeply relevant job matches without the operational overhead.

This is personalization the Blueshift way.

Adaptive. Intelligent. Scalable.

And ready for whatever comes next.

Subscribe to our newsletter for new resources, updates, and more.
Thank you