Profile

LinkedIn Email Github

Summary

ML/ Applied AI Engineer specialized in Natural Language Processing (Traditional NLP, Evaluation of LLM Agents, LLM Fine-tuning) and MLOps (Packaging and optimizing ML models). Previously, an intensively experienced software engineer with the ability to create technology roadmaps, balancing trade-offs with an understanding of the full stack. An effective communicator who harmonizes with stakeholders by addressing concerns and coming up with combined ideas.

Experience

Software Engineer, Ringle, Seoul South Korea

• Apr 2021 ~ Apr 2023
• Ringle Docs Project
  • Product Summary: Collaborative Document Editor replacing Google Docs with needs for custom features(feedback component) and facilitating data utilization.
  • Product ideation, setting up MVP, create technology road-map, On Websocket Backend(NodeJS) keep-alive(retrying), security, editing synchronization features implemented, Infrastructure setup to meet High Availability utilizing Redis as pub/sub queue, MongoDB
• Bite-Sized Feedback Project
  • Reference
  • Product Summary: Feature aiding learners to effectively review what they’ve learned from tutor by breaking down tutor feedback into learning unit, reminding periodically, providing practicing activities
  • Product ideation, setting up MVP, create technology road-map, Implementing features based on Batch Jobs
• NLP Modules
  • Grammar Error Checker
    • Usage: Learning Activity in Bite-Sized Feedback, and measuring component in CAF Project
    • My Contribution: Model Research, Test dataset collection, Packaging and serve as API
      • Due to heavy model running on CPU Instance(cost issue) and requirements for sub-real-time processing, manually implemented batch prediction with request queue pipeline(using ZeroMQ, python).
  • Complete Sentence Checker
    • Usage: Quality Assurance method for detecting insincere feedback sentences from tutors
    • My Contribution: Module Packaging as API and set up serving infrastructure
  • Phrase Extraction
    • Product Summary & Usage: Extracting various english phrases which are worth learning for L2 learners, used for Bite-Sized Feedback Project(as a learning unit)
    • My Contribution: With Python NLP Library(spaCy) ****implemented phrase extraction algorithm, Test data Collection, serve as API
• CAF Project
  • Product Summary: Analyzing L2 performance and proficiency based on Transcribed text
  • Packaging Resource-demanding Natural Language Processing based pipeline
  • Event Driven Architecture with AWS Event Bridge and Kubernetes Job
  • GPU Cluster Setup and Auto Scaling(AWS Spot Instance) for Cost Optimization

Software Engineer, Wordbricks, Seoul South Korea

• Apr 2023 ~ October 2023
• Integrated LLM Management
  • Design interface for integrating LLMs from multiple providers
  • Design VectorDB components for RAG
  • Context-specific LLM parameter optimization and serving

Applied AI Engineer (Intern), Sendbird, Seoul South Korea

• December 2025 ~
• Voice AI Agent
  • Drafting and fast proving integration of Spoken Dialogue Models
  • Design latency optimization module for traditional voice agent pipeline
• LLM Finetuning
  • Design modules for evaluating LLMs on Customer Domain(Policy)
  • Built synthetic dataset for SFT training and filter dataset with train utility analysis
• User simulation research (Ongoing)
  • Design user simulator that mocks real customer utterances
  • Evaluation metric to measure user simulator performance / customer-facing agent