.png)
Anyone who runs a course catalog has watched this happen. A learner needs a specific course, types their best guess at the title, and comes up empty. The course exists, but they just didn't use the words the system was waiting for. This article shows why this issue happens and what you can do about it.
The cost of a guessing game
Keyword search puts the burden on the learner to name things exactly the way the catalog does. Miss the phrasing and the result is a blank page, even when the right content is one synonym away. Someone searching "onboarding" won't find a course filed under "new employee orientation," and most learners won't try a third or fourth phrasing before giving up.
The cost lands in two places.
- Enrollments that should have happened don't, because the learner walked away.
- An underused catalog you spent real time and budget building.
Across a full member base over a year, that's a lot of quiet, hard-to-see missed value.
Describe what you need, in plain language
The intelligent search coming to Cadmium products works from the other direction. Instead of guessing keywords, a learner describes what they actually want in plain language and gets back relevant, permitted results. It's built to read for intent rather than exact wording, so "onboarding" and "new employee orientation" land on the same course.
Each result also comes with a plain reason it matched. The learner can see why a course surfaced and judge whether it fits, and an admin reviewing the catalog can see the same reasoning rather than wondering what the algorithm decided.
A concrete example
Picture a learner typing this into the search box:
"Compliance courses I can finish this week that count for CE"
That's three conditions bundled into one sentence, and keyword search would choke on it. The intelligent search is designed to read it the way a person would and return real options that satisfy all three:
- On topic, because the course is actually about compliance, not just a title that happens to contain the word.
- Finish-able this week, based on the course's length and format.
- Credit-bearing, because it carries CE the learner can count.
This cuts down the chances a learner will receive results hat technically matched a word but didn’t fulfill the three conditions.
Built to be trusted, not just clever
A search that reads plain language is only useful if a provider can trust what it returns. Two things anchor that:
- It respects entitlements. The search only surfaces content a given learner is already allowed to see, so it never exposes a course, or a whole program, to someone without access.
- It reflects the records. When it presents CE, it draws on verified information rather than guessing at credit, so a learner is never told a course counts when the records don't support that.
What the first version does, and what it doesn't
The first version that is currently in development focuses on one job: finding the right content. It focuses on connecting a learner to courses that genuinely fit their needs rather than generating answers or writing summaries on the fly. That focus is deliberate, because in a catalog tied to credit and compliance, returning the right real course matters far more than sounding clever about it.
Elevate will get the intelligent search first, and other Cadmium products follow on the same foundation the rest of our AI is built on. More capability will come once the first version proves out.
To talk through what it could look like in your catalog once it's ready, speak to an expert here, or get the bigger picture in Why Cadmium Built an AI Platform, Not an AI Feature.
