Skip to content
BIO:ON Insight
Yaehyun Lee

About

Yaehyun Lee

MSc student, Biotechnology & Pharmaceutical Management, UCL · I write BIO:ON

BIO:ON is the newsletter I write each week while learning about the biotech and pharma industry. I read the original global coverage and share what I've understood through a Korean lens and a career lens. I'm still learning, so where expert judgement is needed I draw on published industry sources and public interviews with people in the roles, always cited.

For recruiters

I'm looking for graduate programmes and internships starting in 2027.

Target roles
Pharma supply chain & operations, regulatory affairs, commercial & market access
Available from
2027 (after completing my MSc at UCL)
Location
United Kingdom · South Korea
Languages
Korean · English (native/bilingual)

Highlights

  • Excellence Award, AI & Big Data Convergence Competition (Army Signal School, Ministry of National Defense): a pharmaceutical demand forecasting and inventory optimisation concept, selected from around 200 finalist teams (team of 4)
  • Medic, Republic of Korea Army: revised the unit's SOPs for inspecting and managing medicines and medical supplies, adopted unit-wide
  • BIO:ON Insight: writing a weekly Korean/English biotech and pharma newsletter with cited sources since June 2026
  • BSc Pharmacology, UCL (2:1); elected Academic Representative for the pharmacology cohort (2021–2023)

Education

  • Sep 2026 – Sep 2027

    MSc Biotechnology & Pharmaceutical Management

    University College London (UCL)

    Expected
  • Jun 2024

    BSc Pharmacology

    University College London (UCL)

    Graduated
BIO:ON Insight

A media project built with AI

BIO:ON runs on an AI-assisted workflow from research to publishing. This site itself was designed and coded with Claude Code.

  1. 1

    Research & translation

    Global press releases are gathered and machine-translated first, then checked line by line against the originals.

  2. 2

    Newsletter production

    An email-safe HTML template co-designed with AI keeps the structure fixed while the content changes weekly.

  3. 3

    Card news & reels

    A pipeline turns each issue into an Instagram carousel automatically.

  4. 4

    This website

    Built from scratch in Next.js — bilingual, with automatic archive import, and AI translation, summaries and a chatbot on the way.

21 issues published (KO/EN)