$ whoami

Tung

Software Developer — ML, Web & Geospatial

web applications · ML/geospatial pipelines · developer tooling

I build web and geospatial software, from React/TypeScript interfaces to .NET desktop GIS and LiDAR/ML pipelines. I keep the code readable and the data correct, especially when someone else has to run the tool later.

focus · 4 areas open to new work
PUBLIC REPOS WEB + SYSTEMS TESTS + REVIEW

01. Featured Projects

Four public projects — what I built and where I contributed.

02. About Me

How I work through practical software problems, from the interface out to data and delivery checks.

I like practical software problems — especially when maps, LiDAR, or a daily workflow needs a tool someone can actually run.

I hop between the UI, the data, and the release checks when a feature needs all three. That's usually how the web apps, geospatial pipelines, and desktop tools I work on fit together.

  • Correct spatial data — CRS, units, and labels tend to drift if you don't check them end to end.
  • UI, data, and tooling — When a feature needs more than the front end, I fix the data path and the supporting script too.
  • Checks before handoff — I write tests and release steps so "works on my machine" isn't the final state.
Portrait of Tung
software developerHANOI / VN

03. Experience & Education

Geospatial work in desktop software, data processing, testing, and release automation.

04. Project Index

Smaller public tools built to solve a specific problem or automate a repeated task.

More on GitHub →

05. Technical Skills

Rough groups for the work I do most, and the tools I use to ship it.

06. GitHub Activity

A 3D snapshot of my GitHub activity over the past year.

// contributions, visualized in 3D

Use the arrow keys to inspect days, Home or End to jump, and Escape to return to the latest day.

Waiting for activity data…
less more

07. Working with AI

How I plan, build, and verify software with AI help.

// how I work

I use AI for research and boring edits. I still choose the design and what gets merged.

I use AI to read code, compare options, and draft small changes. I review the diff, run the checks that matter, and keep only what I'm happy with.

Working with AI I review before I keep it
  1. Planscope and what done means
  2. Researchcodebase context and constraints
  3. Buildsmall changes that fit the existing code
  4. Reviewtests, behavior, and edge cases
  5. Delivertest results and clear handoff notes

I still decide what gets merged.

Optional 3D sketch of the roles — not a real multi-agent product.

08. Contact

I’m open to software engineering roles and projects in web, ML, geospatial, or developer tools.
If you’re building something in that space, I’d be happy to hear from you.

Say Hello