Research Scientist Intern Mar 2026 – Jun 2026
Krafton AI, Seoul
Contributed to Raon-Speech, a 9B-parameter speech language model achieving state-of-the-art performance across 42 English–Korean speech benchmarks. Served on the evaluation team: contributed to Korean benchmark data construction, introduced and improved the LLM-based judge system for baseline evaluation, and conducted QA testing. Acknowledged in the technical report.