← Back to EngineeringProjects
Production systems, experiments, and tools. What I've built and what I learned.
AI Systems & Automation Consulting
2024-PresentBuilding AI-powered automation systems for enterprises. Focus on practical ROI over hype.
Impact:
- • $2M+ in operational savings
- • 1,000+ users across client organizations
- • 15-30 hours/week eliminated per team
Tech Stack:
TypeScript, Next.js, Python, Anthropic Claude API, OpenAI, vector databases, event-driven architecture
Key Learnings:
- → Users care about outcomes, not AI magic
- → Reliability beats features in enterprise
- → Change management harder than technical implementation
- → Event-driven architecture scales better than RPC for AI systems
Study Forge
2024-PresentPersonal knowledge management system optimized for hyperconnected cognition. Knowledge graphs instead of linear notes.
Tech Stack:
Obsidian, MDX, graph databases, custom tooling
Key Features:
- • Bidirectional links between concepts
- • Module-based learning (non-linear)
- • Pattern extraction from multiple sources
- • Cross-domain connection discovery
Philosophy:
Built for minds that process multiple streams simultaneously. Ma (間) - strategic whitespace. Kaizen (改善) - continuous refinement.
Distributed Consensus Playground
PlannedInteractive visualization of Raft, Paxos, and other consensus algorithms. Make distributed systems intuitive through animation.
Coming Q2 2025
Neural Architecture Experiments
PlannedBuilding transformers from scratch. Exploring attention mechanisms. Understanding AI by implementing it.
Coming Q1 2025
View production work and case studies at /projects