4 Acts. One through-line.
Every chapter follows the same pattern: understand the system deeply, then build something that makes it work better.
Where I've Shipped.
Click each role to see the full breakdown — the problem, my role, what I built, and the impact.
The automotive aftermarket catalog ran on scattered spreadsheets, manual re-entry across Amazon VC / AAP / WHI, and no consistent data model. SKU–MSS–ASIN mappings were fragile. Amazon rejections were frequent. Promotion cycles required heroic effort.
Took ownership of the full lifecycle. Designed a relational ERD spanning SKU, MSS, ASIN, Fitment (ACES), and Attributes (PIES). Built automation in Python, VBA, and UiPath. Embedded a GenAI validation layer for attribute completeness and fitment accuracy. Coordinated between Amazon VSP, suppliers, sales, and engineering.
10-layer relational model: Core Product (SKU) → Distribution (MSS/PTNO) → Marketplace (Amazon ASIN, AAP, WHI) → Fitment (ACES/PIES) → Media → Pricing & Promotions → Inventory → Returns → AI Validation Log → Automation Tasks.
AI-assisted catalog validation for missing attributes and incorrect fitment. LLM-based error explanation for Amazon/AAP rejections. VBA/RPA + LLM hybrid automation for documentation. Rule-based + LLM-assisted product evaluation prototypes.
Fasoo is an enterprise data security and digital rights management (DRM) company serving global clients across government, finance, and tech. Exposure to real production enterprise software, data handling at scale, and cross-functional engineering teams.
The internship established a concrete understanding of how production software systems work — bridging UCSD coursework to real-world engineering. This hands-on foundation directly enabled the technical depth I brought to NAVER Cloud and HL Mando.
Full ownership of product roadmap and backend engineering sprints. Team Management, Task Management, and Back-End Engineering as Tech Lead. Full stack: Frontend, Backend, DevOps, Data Science/ML, and Stock modeling.
Pitched Betcha's market disruption strategy to investors and industry leaders at STEP-UP 2023 (Newport Beach). Then took it to UKC 2023 in Dallas, TX — one of the largest Korean-American professional conferences globally.
Where Research Meets the Real World.
Global conferences, competitions, and exhibitions — from Las Vegas trade shows to university hackathons.
w/ Hyundai MOBIS
w/ Grinda AI
How I Stay Ahead.
- Big Data Infrastructure — Summer 2021 Program at UCSD ↗
- Python Deep Dive — Data Structures & Algorithms ↗
- Algorithm in Depth — research-level analysis ↗
- HW-SW Interactions × Resource Utilization ↗
- Data-Driven Decision Making framework ↗
- Multi-Modality Project Spec in Medical domain ↗
- Tech × Product Trends — GenAI landscape study ↗
- Gen AI × Business — B2B SaaS landscape analysis ↗
- GAN/VAE architectures deep dive
- LLM quantization & inference efficiency
- Neuromorphic Engineering — on-device AI, chip architecture ↗
- CUDA — GPU parallelism, DL accelerator optimization ↗
- Transformer architecture — attention mechanisms ↗
- OpenCV — computer vision, full ML pipeline ↗
- JetBrains Developer Ecosystem 2025 (24,534 global devs)
- AI Jazz Performance × Human-AI Synergy research
- AI × Art analysis — MoMA "Unsupervised" exhibition
- Automotive aftermarket domain: VIO, ACES/PIES, B2B catalog
Things I've Built.
Building Before the Blueprint.
Two ventures where I tested ideas at the intersection of product thinking, engineering, and real-world markets.
The tools that power the work.
From data architecture and GenAI to automotive data standards and automation.
Yewon Hong
Yewon H. builds systems where AI, data, and product execution converge. My work begins with a simple principle: "The Power of Record-Keeping" — every decision I make starts with understanding the system.
From studying Big Data infrastructure at UCSD to rebuilding automotive aftermarket operations with GenAI, each chapter of my career follows the same pattern: understand the system deeply, then build something that makes it work better.
My differentiator is the ability to sit at the intersection of technical depth and product strategy — translating between systems and stakeholders, between data and decisions.
Writing is exploration, not just expression. I write to understand patterns: in people, data, and the subtle connections between logic and emotion. Sometimes code that tells stories through structure. Other times, essays tracing how language mirrors intelligence — both human and artificial.
→ @yzylife_iz