I am a 15-year-old AI & Full-Stack Engineer focused on high-velocity shipping, GPGPU processing, and resilient agentic architectures.
๐ Global Ranking: Placed 10th globally in the San Francisco Agentic AI Hackathon. ๐ Focus: Building self-healing pipelines and web-based parallel compute environments.
๐๏ธ AI-Dubbing-Tool
A production-ready, multimodal multilingual dubbing pipeline featuring an enterprise-grade resilient voice engine.
- Resilience Pattern: Engineered a 4-tier automated fallback cascade (ElevenLabs โ Coqui/XTTS โ Google Cloud โ gTTS) to handle live API rate limits and zero-downtime failures.
- Core Tech:
faster-whisper, Flask, Multi-Engine TTS, Segmented Temporal Translation Alignment.
A high-performance, web-based 4D physics simulation sandbox utilizing web graphics hardware for parallel processing.
- Optimization: Offloaded heavy particle calculations completely from the CPU to the graphics card using custom GPGPU fragment shaders.
- Core Tech: Three.js, WebGL, GPUComputationRenderer, Custom GLSL Shaders.
- Languages: Python, JavaScript, GLSL (Shader Language), HTML/CSS
- AI/ML Pipelines: Whisper, XTTS, LLM Orchestration, Agentic Simulation Frameworks
- Frontend/Graphics: Three.js, WebGL, Reactive UI Components
- Backend: Flask, Event-Driven Orchestration, API Management
- Discord (socials): @william_josh_tech
- X (Twitter): https://x.com/Josh_builds_itz
Open for remote deep-tech contract work, open-source collaborations, and developer grants.