↑ ↓ to navigate ↵ to open · esc to close
ABILAASH / RESEARCH

Better retrieval. Less compute.

Samsung R&D Institute / 2024

Samsung R&D Institute research internship and recognition

RESEARCH INTERNSHIP / RETRIEVAL-AUGMENTED GENERATION

Problem

Retrieve changing e-commerce offers without repeatedly fine-tuning a language model.

Approach

A memory-efficient offers-retrieval system using retrieval-augmented generation.

Results

The existing project account reports 24× cost efficiency compared with its fine-tuning baseline.

My contribution

Research internship implementing the offers-retrieval system.

01Offer information
02Retrieval
03Grounded response

Project account: Build, Learn, Repeat. Metrics describe the internship evaluation, not a general benchmark.

All research