Open Source Projects
Building tools that explore, demonstrate, and solve real problems in AI engineering
UnaMentis Beta
Open-source voice AI learning platform, and the origin project that started this journey. Provider-agnostic architecture lets you swap STT, TTS, and LLM freely: 9 speech-to-text providers, 8 text-to-speech providers, and 5 LLM providers. iOS and web. Learning through natural conversation, anywhere.
UnaMentis was born from a conviction that AI should adapt to human learners, not the reverse. When the voice interfaces to major AI models stopped meeting my needs for in-depth, hands-free learning conversations, I started building my own solution. Within a week, I realized it was heading somewhere with real value.
Pocket TTS for iOS Stable
Ported Kyutai's Pocket TTS (117M parameter model) from Python/PyTorch to Rust/Candle for native on-device iOS inference. Solved 14 major technical challenges across 5 specialized AI agents, achieving near-identical waveform output to the original model.
OpenClaw Active
Governed multi-agent coding architecture. Constitutional governance for autonomous coding agents with structured oversight and drift prevention, building on the patterns discovered through Agent Vision Team.
Agent Vision Team Stable
Multi-agent collaborative intelligence with governance, born from the challenges of building UnaMentis. Uses Claude Code hooks for automatic architectural enforcement across 6 specialized subagents, solving the drift problem in multi-agent systems.
Solution Explorer Stable
Codebase architecture visualization, born from needing to understand complex codebases at scale. Static analysis across 16+ languages with multi-repo support, rendering interactive architecture diagrams.
Knowledge Distillation Framework Active
Independent work exploring an understanding-based approach to machine intelligence. Covers distillation versus compression, dimensional world models, structural integration, and grounding. Published as both a framework and a Substack article.
edu-voice-ai-eval Stable
Voice AI evaluation framework for systematically assessing speech-to-text, text-to-speech, and conversational AI quality across providers and configurations.