Skip to content
Main Concourse Open For Boarding

I make AI systems easier to ship, operate, and understand.

I work on model-serving infrastructure, developer tooling, and technical writing that helps teams move from prototype to production.

Focus

Production AI

Strength

Clear Interfaces

Approach

Hands-On

Departure Board

Current Work

PLT 01 Production inference systems Serving
PLT 02 Developer tooling and explainers Shipping
PLT 03 Technical writing for reusable systems work Writing

Systems

Triton, vLLM, Ollama, TensorRT, and LLM product workflows.

Output

Projects, explainers, and notes built to be inspected and reused.

See Projects Read Writing Contact on LinkedIn
Production Inference Developer Tooling Frontend Explainers
Platform A Proof of Work

Working systems, documented decisions, and reusable technical output.

A compact view of the portfolio, followed by two projects traced from problem to result.

Contact on LinkedIn
Projects indexed
13

Research prototypes, product experiments, and developer tools.

Local-first tools
3

Browser-local utilities that keep commands and data on device.

Technical articles
12

Notes and tutorials spanning ML foundations and implementation.

Documented case studies
2

Problem-to-result breakdowns backed by working projects.

Case Study 01 Python · Reinforcement Learning

Race Car RL Agent

Problem
Teach an agent to navigate a simulated race track from reward feedback instead of scripted steering rules.
Approach
Model driving as a Deep Q-Learning problem and iterate on state, action, and reward design.
Architecture
PyTorch DQN with replay memory and target-network updates inside an OpenAI Gym simulation.
Result
A working research prototype with a paper that documents the training approach and evaluation.
Case Study 02 TypeScript · React

SSH, SCP & rsync Command Builder

Problem
SSH, SCP, and rsync commands become error-prone when connection details, directions, and flags change.
Approach
Model transfer intent with typed inputs and generate commands from deterministic domain functions.
Architecture
React and TypeScript UI, browser-local saved profiles, and pure command builders covered by Node tests.
Result
One local-first tool produces repeatable connection, upload, download, and sync commands without sending credentials to a server.
Platform B Selected Projects

Four representative builds across research, visualization, and developer tooling.

The project yard keeps the complete archive.

View All Projects
  1. Platform 01
    Project
    Race Car RL Agent
    PythonReinforcement LearningOpenAI Gym

    Developed a Deep Q-Learning agent in a simulated racing environment to autonomously navigate tracks by learning from reward feedback.

    Read Paper
  2. Platform 04
    Project
    D3.js Reference Tutorial
    JavaScriptD3.jsData Visualization

    Interactive web guide demonstrating core D3.js features and visualization techniques.

    Source CodeLive Demo
  3. Platform 11
    Project
    SSH, SCP & rsync Command Builder
    TypeScriptReactSSH

    Browser-local developer tool for assembling repeatable SSH sessions, SCP copies, and rsync transfers with saved profiles.

    Open Tool
  4. Platform 13
    Project
    CSV Quick Look
    TypeScriptReactData Visualization

    Browser-local CSV explorer that profiles column types, stats, and distributions without uploading data anywhere.

    Open Tool
Platform C Latest Dispatches

Recent writing on machine learning, engineering, and experiments in progress.

See All Posts
Stop 01

Galton-Watson Process

In this post, we outline the Galton-Watson Process.

Read Dispatch
Stop 02

Policy Iteration Algorithm in a Nutshell

Policy Iteration with Maze Example.

Read Dispatch
Stop 03

Iterative Policy Evaluation in a Nutshell

Check if an agent is doing the right thing with Policy Evaluation Methods

Read Dispatch
Platform D Learning Route
Active Route

Two active topics, with the full study and build notes kept on the learning route.

Current study is focused on reinforcement learning and high-performance systems.

  1. Stop 01
    In Progress

    Applied Reinforcement Learning

    RL algorithms • Policy gradients

    MarkdownPythonObsidianPyTorch
  2. Stop 02
    In Progress

    High-Performance Systems

    Distributed systems • GPU acceleration

    CUDAPolarsTritonParallelization
View the Full Learning Route