XeFoxis
Data engineering, scraping, prediction, and search
- Usually working on
- social data, ranking, recommendations, prediction
- Location
- remote
About
I work on the unglamorous middle of data systems: collecting social data, cleaning it until it is usable, and building models or search logic on top of it. The useful part is usually not the scrape by itself, but the signal you can get after the noise is gone.
scraping
data cleanup
models
search
Things I do
- Write collectors for profiles, timelines, posts, and graph edges
- Turn rough activity data into features that models can use
- Build ranking logic for search results and recommendation feeds
- Check whether a model is helping, drifting, or just looking impressive
Work Notes
- Scraping: I care about repeatable collection, boring validation, and not letting small page changes quietly ruin a dataset.
- Prediction: I like models that can be explained well enough to debug when the numbers stop making sense.
- Search and recommendations: Retrieval first, ranking second, measurement always. Relevance work gets bad fast without feedback loops.