RetroTech Collective

Inspiration

The question that inspired us was: What happens when you combine the aesthetic charm of retro technology with the power of modern AI?

What it does

RetroTech Collective is a portfolio of three interconnected projects that bridge nostalgic aesthetics with modern AI capabilities:

RetroBrain: ML Time Machine - A 3-era machine learning comparison engine that trains models from the 1980s, 2000s, and 2020s on identical datasets.
PixelPalette: Intelligent Pixel Art Editor - A retro-style pixel art editor with intelligent upscaling and auto-shading algorithms.
ElevenLabs TTS: AI Voice Synthesis - A text-to-speech interface using ElevenLabs API for natural, expressive voice generation.

How we built it

RetroBrain Development
- Built modular ML pipeline with era-specific model classes
- Integrated Gemini API with prompt engineering
- Created retro terminal UI with smooth era morphing
- Implemented "AI explains AI" feature
PixelPalette Development
- Built canvas-based drawing engine from scratch
- Implemented custom upscaling algorithm with anti-aliasing
- Added retro color palettes (NES, PICO-8, Game Boy)
- Created save/load system with JSON
ElevenLabs TTS Development
- Designed simple, intuitive interface
- Integrated ElevenLabs API with error handling
- Added audio playback and download functionality
- Implemented voice selection and settings
Eco-Byte Farm Simulator
- Frontend: HTML5 Canvas grid map with tile-based rendering
- Retro UI: 16-bit sprites, pixel fonts, and chiptune audio
- Backend: Optional Flask or Node server to fetch real-time sustainability data
- Game Logic: JavaScript resource management engine
- Dynamic Environment: Pollution, weather, growth cycles all tied to player choices
- Data Input: Real APIs or user-entered carbon footprint habits
Pongatron(Quantum Pong)
- Quantum Engine: JavaScript Pong with superposition and collapse logic
- Retro Renderer: CRT-style vector graphics with 8-bit effects
- QRNG Support: Optional Qiskit/QRNG randomness for ball behavior
- Cross-Platform: Runs in browser or desktop with classic arcade controls

Challenges we ran into

API Rate Limiting & Reliability - Gemini and ElevenLabs APIs have rate limits that could break during demos Solution: Implemented fallback modes with mock responses, request queuing, and error handling

Accomplishments that we're proud of

Complete Portfolio of Three Projects - Successfully built and integrated three distinct projects under unified theme Unique Concept - First portfolio we know of that combines retro aesthetics with modern AI across ML, art, and voice synthesis

What we learned

Deepened understanding of ML evolution from statistical methods to neural networks
Gained experience with multiple AI APIs (Gemini, ElevenLabs)
Learned to handle rate limiting, error handling, and fallback modes