Recommendation Modeling Manager from Advertising/Recommendation Modeling (formerly Machine Intelligence Team)1. New development and advancement of Amazon’s personalized recommendation model (2022. 02 ~ 2022. 10)→ When launching the Amazon personalized recommendation model, sCTR increased by 51%, sCVR by 57%, payment contribution transaction amount by 75%, and number of ordered products by 57% compared to the existing global recommendation model (popular products by category).2. New development and advancement of retail box recommendation model (2022. 12 ~ 2023. 05)→ Reduced courier payment costs, currently all WMS centers (15 out of 19 registered, the remaining 4 test centers) are using the box recommendation model3. Response to product structure improvement and advancement of VT (View Together) recommendation model (2023. 04 ~ 2024. 02)→ As a result of offline evaluation, NDCG@3 increased by 8.03% and NDCG@10 increased by 8.33%4. Efficiency and advancement of repeat purchase recommendation model (2022. 06 ~ 2022. 09)→ 46-56% shorter than existing deployment time, hit count (whether purchase was actually made within the recommended period) increased 4-5 times

Face Recognition (1:1, 1:N) model development and optimization→ Compared to the previous model, performance improved from 98.1% to 99.37%, and speed improved from 1.2 seconds to 0.2 seconds