Arsenal

A living inventory of things I actively use and depend on such as skills, programming languages, certifications, software, hardware, and services.

Tech stack

My go-to languages and frameworks that I use for work and passion.

~/walter-teng/stack

$ stack --list

# Languages

Python My daily driver for AI work, automation, and just getting things done
JavaScriptMy go-to for web apps, Chrome extensions etc
Java Used extensively in my previous roles building microservices for enterprise applications

# AI / Inference

vLLM Preferred LLM inference engine optimized for different hardware. Occasional Ollama use for lightweight local deployment

# Data Engineering

PySpark Spent a good amount of time building and optimizing large-scale data pipelines to process data for LLM training
Airflow I usually paired it with PySpark to automate data pipelines

# Infrastructure

Docker I usually containerize my work; Docker Compose helps me spin up POCs quickly
KubernetesA big part of my current work includes deploying LLM inference at scale reliably in Kubernetes clusters
Terraform My go-to tool for keeping multi-cloud infrastructure consistent and manageable
Slurm Used for scheduling LLM training and distributed workloads across GPU clusters
Ansible Used from bootstrapping my developer machine to configuring GPU clusters

# Observability

PrometheusMetrics collection for monitoring infrastructure, GPUs and applications
Grafana Dashboards for tracking system health, metrics and performance
Langfuse Observability layer for LLM applications

# Frontend