RESEARCH PROJECT
BATCHER: Cost-Effective LLM Inference for Entity Resolution
BATCHER: Cost-Effective LLM Inference for Entity Resolution
Project Lead; First Author · ICDE 2024
- BATCHER: Developed coverage-based demonstration selection and token-aware batch allocation; batch prompting reduced API cost by 4–7× versus standard prompting across eight ER datasets.
fmh1art/BatchER — 12 GitHub stars, checked 2026-09-29.