Negros PowerWatch
Community-focused power outage monitoring platform built for people in Negros Oriental, turning fragmented outage sources into useful public-facing tracking.
- Monitoring
- API
- Maps
- Reports
- Cloud
AI-Assisted Developer • Software Builder • AI Tooling Enthusiast
I build software by combining my own ideas, problem-solving, testing, and system design with AI coding tools and agent harnesses.
My work is largely project-driven. I learn by building real systems, breaking them, debugging them, improving them, and figuring out how different pieces fit together. I am particularly interested in AI agents, developer tools, automation, local AI, web applications, and software that can be useful outside of a demo.
Selected work that demonstrates how I approach building practical systems.
Community-focused power outage monitoring platform built for people in Negros Oriental, turning fragmented outage sources into useful public-facing tracking.
AI coding-agent project exploring reasoning loops, multi-phase execution, provider routing, tool use, memory, verification, and repair.
Rust/Tauri desktop launcher with a terminal-oriented Ratatui interface, designed around instance management and game launcher workflows.
Persistent memory and knowledge infrastructure combining SQLite, FTS5, WAL, and an Obsidian-based environment for cross-session agent context.
Centralized skills system for discovery, relevance ranking, loading, usage, and verification across coding workflows and projects.
Ongoing experiments in automation, local AI, APIs, testing, CLI/TUI tooling, and deployment pipelines using modern development stacks.
Python for AI apps, CLIs, APIs, automation, and testing. Rust for desktop and systems tooling. JavaScript for web apps and APIs. Java in the Minecraft/Fabric ecosystem.
AI coding agents, provider integration, local LLMs, model routing, quantization experiments, AI memory, agent tools, prompt workflows, and verification/repair loops.
Git/GitHub workflows, automated testing, debugging, CLI/TUI development, API integration, Cloudflare Pages, and application packaging.
AI agents, developer tools, automation, local AI, web applications, persistent memory, knowledge systems, and practical software beyond demos.
I am interested in making AI tooling accessible to people without expensive hardware or large API budgets. That means exploring free, local, BYOK, and capability-based routing.
Systems where AI can understand a task, inspect its environment, use tools, modify software, remember context, test its work, detect failures, repair problems, and keep improving.
AI tooling is not a replacement for my role. I use it as a development partner and coding agent.
Start with something I want to build or a problem I want to solve.
Provide requirements, constraints, existing files, expected behavior, and things that should not change.
Inspect repos, write code, modify files, run commands, and work through implementation tasks.
Run tests, inspect output, reproduce problems, and iterate on what the system should actually do.