Skip to content
View Gaan21's full-sized avatar

Highlights

  • Pro

Block or report Gaan21

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Gaan21/README.md

Luciano Plaza

Product Engineer · Full-Stack Software Engineer · AI-Native Development

I build software end to end, from understanding the product problem and shaping the architecture to implementation, delivery, and iteration.

A large part of my current engineering workflow uses AI agents. I focus on the system around them: scoped tasks, project knowledge, static analysis, automated tests, CI, review, and feedback loops that turn recurring failures into reusable constraints.

Selected work

The case study documents engineering decisions behind a deployed SaaS product I build independently: offline-first synchronization, multi-tenancy, authorization, auditability, compliance-informed product design, architecture trade-offs, and AI-assisted delivery. The product source remains private; the case study shares the decisions and evidence that can be exposed safely.

An evolving, visual explanation of a personal AI agent factory: scoped tasks, isolated execution, automated checks, independent review, and human acceptance. It explains the workflow and its boundaries in plain language, distinguishing verified behavior from capabilities still under development.

Executable static-analysis rules derived from recurring architecture and correctness failures. A real timezone defect led to root-cause analysis, three lint rules, automated tests, and CI checks designed to catch future occurrences earlier.

Tooling, in development, for delivering versioned engineering guidance to coding agents. A project selects guidance bundles from Git repositories, pins the exact version, and connects them to Claude Code and OpenCode, so every session starts from the same written guidance. Shared bundles live in agent-guides. Apache-2.0; no stable release yet.

Technical range

C# / .NET · Flutter / Dart · TypeScript / Next.js · PostgreSQL / SQL Server · Supabase · Linux · CI/CD

Links

Pinned Loading

  1. worktracker-case-study worktracker-case-study Public

    Engineering case study for WorkTracker: offline-first systems, multi-tenancy, security, product decisions, and AI-assisted delivery.

  2. agent-factory-case-study agent-factory-case-study Public

    A visual case study of a personal AI agent factory: design, verification, and evolution.

  3. agent-guides-cli agent-guides-cli Public

    Neutral Git-sourced guidance preparation tooling

    Python

  4. dart-architecture-lints dart-architecture-lints Public

    Static-analysis rules for Dart and Flutter covering architecture, timezone correctness, async, lifecycle, and logging.

    Dart

  5. agent-guides agent-guides Public

    Common authored guidance bundles for agent-guides-cli