Cupertino, CA · PhD

Sanghoon Jun

Software Engineer, Machine Learning RecSys

Search · Ranking · Recommendation & Personalization at Scale

EmailLinkedInGoogle Scholar

Ranking & Personalization Leadership

12+ years building and shipping recommendation, ranking, and retrieval systems for profitable, user-facing services — turning business goals into production ML that balances relevance, efficiency, and cost constraints, backed by A/B testing and real-time inference.

LLM Architecture & Workflows

Built agentic LLM workflows and embedding-based search pipelines for recommendation, knowledge-base, and support use cases.

Model Engineering & Deployment

Fine-tuned and deployed domain-specific LLMs for classification, embeddings, and intelligent search.

Technical Lead & Research-to-Product

Drive roadmap and system design across cross-functional teams; translate ambiguous business problems into ML; mentor engineers and data scientists; publications with 1,900+ citations across ML domains.

Prior published researcher — 2,500+ citations across 10 journal publications and 3 patents. See the full record on Google Scholar.

Experience

Senior AI/ML Engineer / Director of Technology · Forkable

2019 – Present

Own the technical direction of a profitable production service, partnering cross-functionally with product and business stakeholders to design, build, and ship ML systems end-to-end and to raise the team's software-engineering practices.

Email me for details.

ML Engineer / Data Scientist · Forkable

2018 – 2019

Owned all data produced daily by the service; deployed models and offline experiments to improve customer experience and inform the team.

Email me for details.

Backend Team Leader / Senior Researcher · Seerslab, Co (YC S16)

2017 – 2018

Led a back-end team and an ML task force; built AWS/Azure cloud systems supporting a mobile app with 100k+ daily active users and improved core service algorithms with machine learning.

  • Improved the company's face-tracking engine with deep learning, data augmentation, and predictive programming; sold the SDK to application and hardware companies
  • Built the content-management and content-delivery services for the mobile application
  • Designed and launched serverless services for 100k DAU with zero server failures

Postdoctoral Research Fellow · Asan Medical Center

2015 – 2016

Managed a computer-aided-diagnosis team building disease-quantification and similar-case retrieval systems for clinical support.

  • Classified lung diseases with deep learning, +5% (91% → 96%) over shallow ML in pattern classification
  • Designed a content-based retrieval system using image pattern recognition and machine learning
  • Advanced imaging, segmentation, and prediction with ML across brain, colon, kidney, heart, and lung diseases

Academic Research Scientist · Korea University

2008 – 2015

Designed and completed research on music information retrieval and music recommendation systems, spanning audio signal processing, large-scale information analytics, and machine learning.

  • Built cover-song identification and similar-song retrieval at 85% accuracy
  • Built a content-based music recommender from user emotion and preferences, 84% user satisfaction
  • Built a collaborative music recommender from social relations and trends over 4.6M analyzed tweets
  • Efficient music-audio representation: mood-sequence generation improved matching accuracy +14% (70% → 84% vs. segment-based); music-structure analysis at 72% segmentation accuracy
  • Music emotion recognition via fuzzy inference, 12% error between user and recognized emotion state

Education

Ph.D. in Electronics, Electrical and Computer Engineering — Korea University

2015

M.S. in Electronics, Electrical and Computer Engineering — Korea University

2010

B.S. in Electrical Engineering — Korea University

2008

Skills

Machine Learning & AI

PythonPyTorchLLM fine-tuning & embeddingsRecommendation / Retrieval / Ranking systems

Systems & Infrastructure

Real-time inference & model servingLarge-scale data pipelinesDistributed systemsAWSSQLMLOps / model deployment

Product & Experimentation

A/B testing & offline metricsOR-ToolsRuby on Rails

Publications

View full profile on Google Scholar →