Jatin0Jain/Redrob-Candidate-Ranker

chore: regenerate team_submission.csv using restored pipeline

revert: rollback to pre-XGBoost state (commit e9d97dc) with original rule-based heuristic

remove progress bar to fix stuck button on repeated clicks

fix demo: display pre-computed 100k ranking results instead of live re-run

add reasoning block per candidate using team_submission.csv as primary source

fix: add missing candidate_embeddings.npy and guard lazy-load against partial failures

upgrade demo to run full 100k ranking pipeline with lazy-loaded artifacts

replace sample candidates with real top-ranked AI/ML profiles for accurate demo

fix demo scoring: compute real sentence-transformer embeddings for semantic_score feature

add sample_candidates.json and sync app.py for demo to work correctly

upgrade: XGBoost ML ranker + consistent demo app

Upload requirements.txt

Upload requirements.txt

build gradio ui

add cached llm reasonings

core ranking logic

upload full 100k candidate dataset and parquet

add constants and deps

initial space setup