Tell us about yourself to find your perfect match in London.
Finding your matches...
Analyzing compatibility using AI
Top Matches For You
Check Specific Compatability
Curious about someone specific locally?
Compatability Result
Manage Dataset
Initialize the recommender system with a new CSV dataset.
Drag & Drop CSV File
or click to browse
Selected File:
Requirements: CSV must contain name, interests,
area, and postcode columns.
Dataset Preview (Top 100)
A sneak peek into the currently loaded profiles.
Loading data...
Name
Area
Interests
No data available. Please upload a dataset.
How It Works
Understanding the AI-powered recommendation logic.
1. Profile Parsing
We use an LLM to extract structured interests (likes/dislikes) from
your free-text input.
2. Vector Embeddings
Your interests are converted into mathematical vectors (lists of numbers) using
OpenAI's text-embedding-3-small model. This allows us to understand the
semantic meaning of your hobbies.
3. Similarity Search
We calculate the Cosine Similarity between your vectors and everyone else's.
LL (Likes-Likes): Do you enjoy similar things? (High weight)
DD (Dislikes-Dislikes): Do you dislike the same things? (Small bonus)
LD/DL (Clashes): Does one person hate what the other loves? (Penalty)
4. Geospatial Filtering
We use postcodes to calculate the Haversine distance between you and
potential matches. The final score is a weighted combination of Interest Match (80%) and
Proximity (20%).