Dongjin Kim

Toronto, ON | Permanent Resident | 343-368-2338 | [email protected] | linkedin.com/in/dongjink | github.com/dkim3005 | djkimlab.com
Technical Skills
Languages: Python, C++, JavaScript, Bash, SQL, MATLAB
ML & AI: PyTorch, TensorFlow, scikit-learn, OpenCV, CLIP, DenseNet, OpenAI & Claude APIs, LLM tool-use & agentic pipelines, RAG, NLP
Systems & Cloud: FastAPI, WebSocket, Docker, Kubernetes (CKA certified), Jenkins, Git, Azure, Grafana, Prometheus, Cloudflare
Experience
Consulate General of the Republic of Korea in Toronto Toronto, ON
Administrative Officer — Technical Systems, Political & Economic Affairs Dec 2025 – Present
  • Built a Python pipeline ingesting ∼15 government and media sources, classifying articles by NLP topic and generating LLM-summarized digests — runs 5×/day with monitoring, logging, and retry logic.
  • Built a searchable internal dashboard consolidating stakeholder profiles, company activity, and archived briefings into one queryable store.
  • Cut a 2.5-hour manual briefing process to a 20-minute review workflow; automated Telegram digest distribution, secured with Cloudflare Zero Trust.
Luxoft — Hyundai Motors Infotainment Platform Seongnam, South Korea
DevOps Engineer Feb 2022 – Dec 2022
  • Owned the CI/CD pipeline for a vehicle infotainment platform — Jenkins builds, Docker images, and deployments across 20+ Kubernetes nodes.
  • Built 10+ Grafana dashboards and Bash-based alerting; caught an architecture misconfiguration on a 128 GB production server before impact.
CSI Vision Suwon, South Korea
Software Application Engineer Mar 2021 – Jul 2021
  • Built a MATLAB algorithm reconstructing 3D molecular structures from 2D cross-sectional HDF5 data for the Korea Institute of Materials Science ($20K contract), with interpolation to handle missing measurements.
Projects
Investment Operations Suite ops.djkimlab.com

Live back-office platform — five modules (reconciliation and break detection, NAV/performance reporting, data quality monitoring, trade-confirmation extraction, double-entry accounting) over a market simulation anchored to real FRED data. FastAPI, Next.js.

Radiology AI radiology.djkimlab.com

Chest X-ray platform over the NIH ChestX-ray14 corpus — DenseNet121 multi-label triage with FHIR context, Class Activation Map explainability, and CLIP semantic image search on one shared model.

Autonomous AI Agent agent.djkimlab.com

A tool-using agent that plans, acts, observes, and self-corrects until a task is solved; tools include a restricted, AST-validated Python sandbox. Built on LLM tool-calling.

More live at djkimlab.com — Reinforcement Learning Lab (DQN, MCTS, AlphaZero-style self-play), Audio Intelligence, Quant Trading Platform, and a RAG chatbot. Georgia Tech coursework: Deep Learning (CS 7643), Machine Learning (CS 7641), ML for Trading (CS 7646).

Education
Georgia Institute of Technology Atlanta, GA / Online
M.S. in Computer Science (Artificial Intelligence specialization) — May 2026
Korea National Open University — B.S. in Computer Science 2022
Kyung Hee University — B.B.A. in Business Administration 2017
Certifications

Certified Kubernetes Administrator (CKA), 2023  ·  TensorFlow Developer Certificate, 2021