Case studies
One page per project: the problem it was built for, how it works, and where it stands today. Most of the code is private, so each page says plainly what is live, what runs locally and what has been retired.
Scorpia Monitor & Quantum Processor — Operations Center with a Voice AI Agent
An operations center for every Scorpia service, drawn as a live 3D field, with a voice AI agent at its core. Tap, speak, and Scorpia checks the systems, answers out loud, and speaks up on its own when something breaks.
Scorpia Stream & Media Discovery
A cinema-grade movie and TV series discovery web application mounted at /stream. Features a widescreen spotlight carousel, trending shelves, and multi-criteria filtering across genres, release years, and ratings.
Scorpia CRM & Pipeline Desk
A multi-tenant deal and recruitment pipeline manager built to handle leads, candidate stages, interaction logs, and credit-backed background jobs in one unified workspace.
Scorpia Batch Image Studio SaaS
Catalog-scale generative AI application that applies one visual instruction across whole product shoots and eCommerce catalogs, orchestrating FLUX, Qwen, and Gemini model endpoints with fail-safe credit accounting.
Opportunities Myanmar & CV Studio
A social-impact web portal cataloging verified international scholarships, academic pathways, and work visas for Myanmar students. Includes an interactive in-browser CV Studio to compose ATS-standard resumes with print-ready PDF export.
Minutes AI — Meeting Notes from a Recording
Records a meeting from the microphone in the browser, transcribes it live, labels who is speaking, and turns the transcript into structured notes: a summary, topics, decisions and action items with owners.
Scorpia Personal AI Coding Desk
A developer assistant workbench combining local offline GGUF models with cloud LLMs, equipped with terminal execution, workspace file tools, git operations, and push-to-talk speech input.
Generative AI Image & Video Lab
Self-hosted image generation behind the Scorpia studio: Stable Diffusion through the Forge WebUI API, the Realistic Vision checkpoint with custom LoRA adapters for consistent photoreal product and portrait shots, a Qwen-Image-Edit / FLUX GPU server, and ControlNet pose guides for AI video edits.
Jarvis — Offline Voice Coding Assistant
Say "Hey Jarvis", speak a request, and it is transcribed and handed to Aider wired to a local LLM as a one-shot coding command. Wake word, speech-to-text and the model all run on the machine, so nothing leaves it.
Coding Agent Server on a Free GPU
Runs Qwen2.5-Coder-7B on a Colab GPU with llama.cpp and exposes it through a token-protected private tunnel, so local coding tools (Aider, Continue, Cline) get a fast open-weight model without paid APIs.
Now Building: AI Shorts Pipeline & Games
An automated short-form video pipeline that takes an idea through AI-assisted scripting, generated visuals, FFmpeg editing and publishing, plus my own game projects covering gameplay programming, assets and playtesting.