Overview
What this challenge is about.
Use a pretrained vision-embedding model (CLIP ViT-B/32 or DINOv2-small). Index a catalog of around 1,500 furniture images. Curate a small evaluation set of around 50 user-style photos paired with the 'correct' catalog match (annotator-verified). Compute Recall@5 on the evaluation set. Build a tiny Streamlit demo. Deliver the index, demo, evaluation, and a 3-page memo on what the product team should A/B test next.
The Brief
What you'll do, and what you'll demonstrate.
Prototype an image-similarity search with at least 0.70 Recall@5 on user-style photos and recommend the next A/B test.
Earning criteria — what you'll demonstrate
- Use pretrained vision embeddings for a similarity-search task
- Build a small vector index and query it efficiently
- Evaluate retrieval with Recall@k
- Recommend product-experiment next steps from a working prototype
Program Fit
Where this fits in your program.
Sharpens the same skills your degree expects you to demonstrate.
Skills
Skills you'll demonstrate.
Each one shows up on your verified credential.
Careers
Roles this prepares you for.
Real titles. Real skill bridges. Pick the one closest to your trajectory.
Computer Vision Engineer
Image-embedding-based search is a portfolio-friendly CV engineering project that any consumer-AI or commerce-AI team values.
This challenge sharpens
- image-embeddings
- vision-transformers
- image-search
AI Engineer
Wrapping pretrained models + vector indices as a working demo is the AI-engineer skill that ships features in startups.
This challenge sharpens
- image-search
- ui-prototyping
- image-embeddings
AI Product Manager
Designing the A/B test off a working prototype is the AI PM's craft of turning models into measured product moves.
This challenge sharpens
- evaluation
- ui-prototyping
- image-search