Find your next skill.
Describe what you know and where you're headed. Aptivio retrieves the courses whose reviews actually match your intent — ranked by similarity, not by what's popular.
Coverage
80
Courses indexed
109,776
Reviews scored
12
Categories
7,815
TF-IDF terms
Real numbers from the indexed dataset, not marketing copy.
How it works
Retrieval, not a black box.
- 01
Describe your intent
Type what you already know and where you want to go, in plain language — no keyword search syntax.
- 02
Vectorize the query
The same TF-IDF vectorizer used at training time turns your sentence into a weighted term vector.
- 03
Compare against every review
Cosine similarity scores your query against all 109,776 course reviews in the corpus, not a cached subset.
- 04
Rank by best match per course
Each course keeps only its single best-scoring review, then courses are ranked by that score — no popularity weighting.
Why retrieval
Deterministic, not generative
Same query, same vectors, same ranking — every time. Nothing is hallucinated or paraphrased by a language model.
Every review, every time
No pre-filtered shortlist. Your query is scored against the entire dataset, so a course can't be missed just because it's less common.
Scores you can inspect
Match percentage is relative similarity, shown plainly — not a hidden relevance algorithm tuned for engagement.
From the review corpus
Real excerpts from the same dataset the matching engine searches — not marketing testimonials.
“Python for Data Science exceeded my expectations from the very first module. Working with NumPy arrays and Pandas DataFrames was the highlight of this course.”
— Python for Data Science
“I want to share my detailed thoughts on Machine Learning Fundamentals after completing the full program. Training and evaluating models using scikit-learn was practical.”
— Machine Learning Fundamentals
“I recently completed JavaScript Full Stack Development and it was a fantastic learning experience. Connecting a React frontend to a Node backend finally clicked.”
— JavaScript Full Stack Development
FAQ
Is this powered by an LLM?+
No. Matching is TF-IDF vectorization plus cosine similarity — a retrieval method, not a generative model. Nothing here is written by an AI at request time.
What does the match percentage actually mean?+
It's relative similarity within a single search: the best result is set to 100% and everything else is scored against it. It reflects closeness to what you typed, not a verified accuracy figure.
How many courses are indexed?+
80 courses, drawn from 109,776 individual course reviews used to build the similarity index.
Can I save or bookmark a course?+
Yes — create a free account and bookmark courses from Dashboard or Discover Paths. Your skills profile also generates a personalized Roadmap.
Where does the review data come from?+
A training corpus built for this project, used to construct and evaluate the retrieval index.
Accounts, saved paths, and a skills profile are live.
Create an account to bookmark courses and get a roadmap generated from your own skills.
Create your account