South Korea
💼 Career Interests Currently seeking opportunities in the following areas: - Efficient Model Inference - Code LLM, LLM, and Applied AI - Compiler and Programming Languages - NPU Compiler and Runtime 🛠️ Technical Skill Set - Languages: C/C++, Java, Python, C# - NPU & Compiler: NPU Compiler, Graph Optimization, Quantization, Kernel Programming, ONNX, ONNX Runtime, TVM, MLIR, LLVM - Efficient Inference: vLLM, Chunked Prefill, Prefill Disaggregation, Speculative Decoding - ML Models & Framework: Transformer, BERT, GPT, MoE Models, PyTorch, TensorFlow, Hugging Face - Distributed Computing for LLMs: Data Parallelism, Model Parallelism, Pipeline Parallelism, Expert Parallelism - Databases & Big Data: MySQL, Elasticsearch, MongoDB, BigQuery, Dataflow, Hadoop, Spark - Software Frameworks: Android Framework (AOSP) - Backend Server: Spring Framework, FastAPI, Flask - DevOps & Cloud: Ansible, Kubernetes, Docker, Jenkins, AWS, GCP 🎓 Graduate Research Focus - Programming Languages and Compiler Design - Source Code Analysis - Compiler Construction - Embedded Virtual Machines 🔬 Current R&D Interests - Optimizing AI Model Inference - Integration of DL Frameworks (vLLM, PyTorch) with NPU - NPU Compiler Development - Code LLM and General LLM Research 📚 Previous R&D Experience - NLP for Smart TV Devices - NLU Builder Development - NLP-based Recommendation Systems - Open-domain Chatbot Models and Services - POI Data Management Tool Development - Compiler, Program Analysis: Static/Dynamic Analysis, Secure Coding - Source-level Application Translators (e.g., WIPI to Android) - Embedded Virtual Machine Research - Middleware/Framework and Application Development for Embedded Linux - SoC Verification Software Platform - Android Framework Development (AOSP)
* Affiliation updated following the merger of Sapeon and Rebellions. 1. Developing PyTorch and vLLM hardware plugins optimized for NPUs. 2. Conducting research and development on methodologies for efficient model inference, such as P/D disaggregation, chunked prefill, paged/flash attention, etc.
1. NPU Compiler 2. Graph Optimization 3. Quantization 4. ONNX, ONNXRuntime
1. Big Data Preprocessing: Apache Beam, Dataflow, BigQuery, Hadoop 2. Dialog Model Research using Large-scale Language Models: GPT NEO, Blender, OpenAI GPT-3 3. Chatbot Service Research & Development 4. AI Chatbot BLOONY Service Research & Development: https://bloony.ai 5. Safety Detection API Research: https://demo.tunib.ai 6. Infra Solution Architecture and Technical Supporter: AWS, GCP, CD/CI
1. Opensource NLP Library PORORO Research & Development: https://kakaobrain.github.io/pororo 2. Pre-trained Model Research & Development 3. Chatbot Model and Service Research & Development
1. NLU Builder (NER, Intent Classifier) Server Development 2. NLP based Recommend Algorithm Research and Development 3. Expert/Contents Recommend System Research and Development (Contents Based Filter, Collaborative Filter) 4. Contents Search Platform Development 5. POI Data Management Tool Research and Development 6. Influencer Trends Service Development 7. Test Automation Platform Development Etc - Certification from NVIDIA FUNDAMENTALS OF DEEP LEARNING FOR NATURAL LANGUAGE PROCESSING