A curated set of technical notes, experiments, and architectural sketches across AI, optimization, and complex systems.
This is my working space for ideas in progress — a mix of design decisions, implementation details, and reasoning around systems I’ve built or explored. Some are deep dives into foundational topics; others are quick outlines or in-development research.
| Topic | Description |
|---|---|
llm_eval/ |
Exploring trust, truthfulness, and chain-of-thought integrity in language models |
scheduling/ |
Notes on GA-based scheduling, changeover sequencing, and constraint systems |
forecasting/ |
Demand shaping, residual modeling, and inventory-aware forecasting strategies |
agents/ |
Planning, memory architectures, and prompt engineering for tool-using LLMs |
systems/ |
Observability, integration architecture, and SaaS platform decisions |
cto_memos/ |
Long-form thinking on product tradeoffs, team design, and scale patterns |
I’ve worked across AI systems, manufacturing optimization, forecasting, and infrastructure — often at the intersection of decision-making, automation, and real-world constraints. These notes help document:
- Models and experiments worth revisiting
- Concepts I’ve had to explain repeatedly (to teams, execs, or myself)
- Unsolved or half-solved problems I think about often
- LLMs as reasoning engines, not just text predictors
- AI-native interfaces for operational planning and control
- The hidden complexity of “real-world integration”
- When to trust a model (and how to know if you shouldn’t)
Most folders include:
.mdmemos (designs, ideas, results)- Code sketches or notebooks (lightweight or illustrative)
- Pointers to real implementations (when public)
This is not a polished library — it's closer to a technical journal.
