CTO @ DoozieSoft | Generalist | Systems Thinker | Builder

This GitHub space reflects my thinking surface—experiments, drafts, boundaries, and unfinished work. It is not a portfolio, showcase, or sales channel.

What you see here is how I reason about systems before they are production-ready.

I focus on systems that survive real-world constraints, not demo environments.

• Systems that reduce cognitive load Fewer decisions. Clear flows. Predictable failure modes.

• Tools that force clarity Explicit intent. Named contracts. Visible assumptions.

• AI workflows that respect human intent Assistive by default. Bounded always. Non-autonomous.

• Most software fails due to unclear intent, not bad code.
• AI must be non-autonomous by default—authority requires accountability.
• Discipline beats motivation every single time.
• Money is a byproduct of usefulness, not a primary objective.
• Simplicity compounds; complexity taxes everything downstream.
• Slow, steady execution outlives fast, reactive engineering.

• AI-native software design Designing systems assuming AI assistance exists—but must be constrained.

• Intent → Contract → Code pipelines Making intent explicit before implementation begins.

• DIL (Dooz Intent Language) A human-readable intent layer to reduce LLM hallucination and token waste.

• Agent boundaries & NONOMOUS doctrine Clear lines between suggestion, execution, and approval.

• Brownfield system modernization Improving live systems without breaking trust or continuity.

• Founder-level engineering discipline Systems thinking beyond code: delivery, documentation, accountability.

• Experimental repositories
• Internal tools spun out once they prove useful
• Architecture drafts and working notes
• Readmes that read more like doctrines than docs
• Half-built ideas—intentionally left visible

If something looks unfinished, it probably is—and that’s the point.

• Tutorial fluff
• Trend chasing
• Over-engineered “big tech” abstractions
• Fake scalability narratives
• Autonomous agents making business decisions

If a system cannot explain itself, I’m not interested.

I collaborate selectively on:

• Open-source projects with philosophical and architectural depth
• Tooling for developers, founders, and small teams
• AI systems with clear guardrails, auditability, and human checkpoints

If your idea needs hype to survive, it’s not for me. If it needs clarity, discipline, and patience—it might be.

This GitHub is not a portfolio. It is how I think in public.

If you’re looking for polished outcomes, look at products. If you’re looking for reasoning, boundaries, and direction—you’re in the right place.