Chun Wang

Senior Research Engineer with 6+ years building production-grade ML systems, dynamic pricing models, and TB-scale data pipelines. Currently driving Bandit-based exploration strategies at LINE GO. Expert in Information Retrieval (RAG), Recommender Systems, and end-to-end ML deployment. Champion of the DevOps philosophy, leveraging AI-accelerated workflows to deliver highly observable, low-latency solutions.

Senior Research Engineer at LINE GO (2026-02 – Present)

• Dynamic Pricing & Experimentation: Researching Bandit algorithms to optimize online exploration and dynamic pricing. Designing robust A/B testing frameworks and owning the full ML lifecycle—from algorithmic research to deploying low-latency models in a high-concurrency production environment.

AI Engineer at Voiss (2025-04 – 2025-09)

• Information Retrieval (RAG) & Deployment: Engineered a scalable RAG pipeline using Elasticsearch and vector DBs. Optimized parsing and context retrieval, hitting 93% accuracy for complex queries. Deployed low-latency services on Kubernetes (GKE) to ensure high throughput for streaming applications.

Generative AI Engineer at Answer.AI (2024-10 – 2025-03)

• Content Understanding & Eval Pipelines: Built an intermediate reasoning layer to interpret OCR data, injecting structured context to boost query success rates by 42%. Architected automated LLM evaluation using DeepEval, cutting testing time by 30% while establishing safety guardrails.

MLOps & Data Engineer at TSMC (2022-07 – 2024-10)

• TB-Scale Pipelines & System Reliability: Architected distributed data pipelines on Kubernetes and Airflow processing 2-5 TB of daily logs. Developed unified FastAPI backends and implemented GitOps (ArgoCD) CI/CD, ensuring 99.9% availability for 200+ concurrent mission-critical jobs.

Senior NLU Engineer at Ubestream (2021-04 – 2022-02)

• Platform Scalability: Transformed bespoke NLU/Content Understanding projects into a scalable, modular SaaS architecture, aggressively reducing deployment time from 1 month to 3 days.

AI Developer at Silkrode (2019-07 – 2021-04)

• End-to-End ML System: Trained and deployed BERT models for intent classification (93% F1 score), automating Tier-1 customer support in a production environment.

Master in Network Learning Technology – National Central University (2017 – 2019)

Bachelor in Information Management – National Formosa University (2013 – 2017)