Hi, I'm Hochan Son

Datastore, SRE, DevOps Engineer & ML Practitioner based in Los Angeles, CA.

I build data infrastructure, ML pipelines, and distributed systems. My background spans ADtech, entertainment, and enterprise — from MySpace and Hallmark Labs to Branch.io and ADP, with graduate work at UCLA Trustworthy AI Lab.

Areas of Focus

• Synthetic data generation (using Transformer, VAE, Diffusion models) & privacy-preserving ML
• Legacy Data Ops to SRE & DevOps to scale in the Cloud Native Infra
• Large-scale data/ML pipelines (MLFlow, Kafka, LMDB, distributed training)
• Local LLM inference & serving (CUDA, MLX, RDMA, vLLM, Ollama)
• Large-scale Database engineering (RDBMS, NoSQL, and Distributed SQL)
• CI/CD & containerized for ML workflows (Docker, kubernetes, GitHub Actions)

Tech Stack

• Languages: C, Python, SQL, Go, Bash, Java
• ML/AI: PyTorch, Diffusion, Variational Autoencoder (VAE), vLLM, MLX, MCP, Agents
• Data: Kafka, SQLite3, LMDB, Redis, PostgreSQL, MySQL, MS SQL Server, ProxySQL, Aerospike, MongoDB, OpenSearch, FoundationDB, etcd, Prometheus, memcached
• Observerbility: Cloudwatch, Datadog, Grafana, Log Stash
• Infra: Kubernetes, Docker, HPC (Distributed GPU Training), GCP, AWS
• CI/CD: GitHub Actions, ArgoCD, Gitlab, Jenkins, Drone.io, ECR, GCR, Dockerhub

Education

• UCLA — Master of Applied Statistics Data Science (MASDS)
• University At Buffalo - B.S. Computer Science & Engineering

Publication

SYNTHONY: A Stress-Aware, Intent-Conditioned Agent for Deep Tabular Generative Models Selection, ICLR 2026 - DeLTA, accepted (poster) [https://arxiv.org/pdf/2604.00293]

Connect