Browser-safe workflow lab
Paste rough game notes, inspect how they are chunked, categorized, validated, and assembled into a production map.
Python local AI tool for game production planning
Turn rough game design notes into structured production maps.
AYO Chat Agent v1 is a portfolio-ready software case study: a simple Python desktop app that loads design documents, runs a controlled local Ollama/Phi-3 pipeline, validates responses, categorizes useful details, and outputs a Markdown production map.
Try the production logic
Clear technical boundary
The lab demonstrates the orchestration contract deterministically in the browser. The private local Ollama model and production implementation are not shipped here.
Inspect It Two Ways
Interactive Browser Lab
Operate the public workflow contract directly: intake notes, watch each stage execute, inspect the production map, and review missing decisions.
Recorded Production Build
The walkthroughs show the private local-model application, its architecture, test behavior, and packaged milestone without distributing the executable.
30-Second Reviewer Scan
Problem
Game design notes are messy. Mechanics, lore, abilities, and production tasks get buried in the same document.
Solution
AYO converts design text into categorized production-map sections that are easier for a developer or small team to act on.
Proof
The showcase includes 28 passing tests, an interactive browser lab, UML diagrams, thumbnails, chapters, and milestone videos.
Best Fit
Software development, AI tooling, game tools, technical design, developer support, and instructional technology roles.
What It Solves
Long design documents mix mechanics, story, UI notes, systems, and production concerns. AYO helps turn raw design text into useful development categories.
How It Works
The workflow combines a desktop interface, loader, chunker, local AI provider, response verifier, strategy-based categorizer, and production-map assembler.
Why It Matters
This is controlled AI tooling, not a prompt demo. It shows architecture, validation, tests, packaging, and game production thinking working together.
Architecture Snapshot
Clean technical diagrams are supported by additional generated UML and original exports on the diagrams page.
Reviewer Path
For a fast technical review, start with the final walkthrough, scan the diagrams, then use the testing video and README to confirm engineering discipline.
Evidence Trail
End-to-End Walkthrough
Shows the loader, chunker, verifier, build flow, and final milestone behavior.
Architecture Walkthrough
Shows Strategy Pattern categorization and production-map assembly.
Testing Walkthrough
Shows repeatable tests, validation behavior, and development discipline.
Milestones
Desktop Workflow
Keep the local document-to-production-map experience stable, clear, and easy to test for reviewers and collaborators.
Tray and Model Status
Add background running, local model status, and clearer visibility into when Ollama and Phi-3 are ready.
Guided Setup
Build a first-run setup flow that walks users through Ollama install checks, model pull, and local configuration.
Optional Integrations
Expand into opt-in workstation support for tools like VS Code, Visual Studio, and creator-oriented local workflows.
Portfolio Boundary
- Public page shows architecture, workflow, tests, and milestone evidence.
- Hosted proof is available through Matteo for reviewers who want a live browser experience without local setup.
- The public lab demonstrates the orchestration contract; the private executable remains available only for authorized review.
- Private creative IP, unfinished internal systems, and implementation details stay controlled.
- Earlier course videos may reference the original prototype title. This repo is the MMKPC-owned AYO Chat Agent v1 showcase.