Aug 22 • General discussion
Anyone else drowning in course content right now?
Is this achievable?
Looking to vectorize course content and create personalized roadmaps for each user.
Been thinking about this all weekend after buying my 7th Skool course I'll never finish.
The technical approach I'm exploring:
- Vectorize all course content using OpenAI embeddings
- Store in Pinecone/Supabase vector DB
- AI assessment to understand user's current state
- Generate personalized paths based on semantic similarity + user goals
- Track progress and adapt as they go
- Chunking content effectively (too big = lose nuance, too small = lose context)
- Balancing prerequisite requirements with personalization
- Making assessment feel conversational, not like a test
- Deciding whether to use pure content vectors or add behavioral signals
Made a quick video showing what I've got so far:
My real question:
Is this overengineering the problem?
Like, do we even need AI-powered personalization or would better UX solve the same issue?
Anyone else working on similar vectorization/personalization problems? What approach are you taking?
Specifically curious about:
- How you handle prerequisite dependencies in personalized paths
- Whether you're using behavioral data or just content embeddings
- If anyone's seen actual completion rate improvements from personalization
genuinely trying to figure out if this is worth pursuing further or if I'm solving the wrong problem.
4:38
1
1 comment
Anyone else drowning in course content right now?
skool.com/vibe-coders
Master Vibe Coding in our supportive developer community. Learn AI-assisted coding with fellow coders, from beginners to experts. Level up together!🚀
Leaderboard (30-day)
1
+12
2
+10
3
+9
4
+7
5
+5
Powered by