Notes

A personal knowledge management (PKM) system for professional work in data science, machine learning, and AI. Built in Obsidian.

---

Purpose

This vault is a structured, long-lived knowledge base — not a note dump. It is organized by epistemic layer and abstraction level: each layer answers a distinct question, from why does this work mathematically? through to why are we solving this business problem?

The goal is to accumulate knowledge that compounds over time: timeless theory in foundations, analytical practice in data science, reusable model knowledge in modeling, proven engineering patterns in the engineering layers, domain use cases in applications, and concrete implementations in reference implementations.

---

Structure

notes/
├── 00_meta/                        # Vault governance: conventions, templates, dashboard, glossary
├── 01_foundations/                 # Timeless mathematical and theoretical foundations
├── 02_data_science/                # Analytical reasoning: from data to decisions
├── 03_modeling/                    # Model families, training, and model selection
├── 04_software_engineering/        # Programming languages, system design, tooling
├── 05_ml_engineering/              # ML production lifecycle: data, training, deployment, monitoring
├── 06_ai_engineering/              # Foundation model systems: RAG, fine-tuning, inference, LLMOps
├── 07_applications/                # Domain use cases and business application patterns
├── 08_implementations/   # Cross-layer executable patterns: system patterns and end-to-end examples (concept-specific impl lives with the concept)
├── 09_projects/                    # Time-bound execution instances
│   ├── _active/
│   ├── _completed/
│   └── _experimental/
├── 10_reading/                     # Structured intake: papers, books, articles, reports
├── 11_logs/                        # Operational memory: daily notes, reviews, brain dumps
└── 99_archive/                     # Retired content

---

Layer Guide

Layer	Guiding Question	Examples
01_foundations	Why does this work mathematically?	Calculus, linear algebra, backpropagation theory
02_data_science	How do we reason from data to decisions?	EDA, feature engineering, experiment design, interpretability
03_modeling	Which model should we use and why?	GLMs, neural networks, transformers, evaluation, regularization
04_software_engineering	How do we build robust software?	Python, Go, TypeScript, APIs, databases, testing, DevOps
05_ml_engineering	How do we productionize ML systems?	Feature stores, MLflow, drift detection, serving
06_ai_engineering	How do we operate LLM systems?	RAG, LoRA fine-tuning, quantization, LangSmith
07_applications	Why are we solving this problem?	Fraud detection, churn prediction, document intelligence
08_implementations	How is this cross-cutting executable pattern implemented?	Reusable system patterns, end-to-end reference architectures (concept-specific impl → home layer)
09_projects	What are we executing right now?	Active work, experiments
10_reading	What have we learned from external sources?	Paper notes, book summaries
11_logs	What happened today / this week?	Daily notes, reflections

---

Note Types

Notes follow templates matched to their layer (see 00_meta/templates/):

Template	Layer	Structure
tpl_foundation.md	01_foundations	Statement → Intuition → Mathematical Formulation → Assumptions → Consequences
tpl_proof.md	01_foundations	Statement → Assumptions → Proof → Notes
tpl_data_science.md	02_data_science	Problem Context → Analytical Goal → Data Considerations → Method → Validation
tpl_model.md	03_modeling	Core Idea → Math Formulation → Inductive Bias → Strengths/Weaknesses → Variants
tpl_software_note.md	04_software_engineering	Purpose → Core Concepts → Design Notes → Trade-offs → Implementation Notes
tpl_ml_system.md	05_ml_engineering	Purpose → Architecture → Data/Feature Flow → Operational Considerations → Trade-offs
tpl_ai_system.md	06_ai_engineering	Use Case → Model Strategy → Evaluation → Guardrails → Architecture → Cost/Latency
tpl_application.md	07_applications	Problem → Stakeholders → Domain Context → Inputs/Outputs → Modeling Options → Risks
tpl_reference_implementation.md	08_implementations	Goal → Conceptual Counterpart → Dependencies → Code Pattern → Practical Notes
tpl_end_to_end_example.md	08_implementations	Goal → Problem Setup → Data → Design → Implementation → Evaluation → Extensions
tpl_project_overview.md	09_projects	Goal → Scope → Deliverables → Data → Modeling → Engineering → Timeline
tpl_reading_note.md	10_reading	Source → Type → Why It Matters → Key Ideas → Relevance to Vault

All notes carry frontmatter: layer, type, status (seed / growing / stable), tags, created.

---

Conventions

Full conventions are in 00_meta/conventions.md. Key rules:

• File names: lowercase with underscores (chain_rule.md, not ChainRule.md)
• Top-level structure is stable — changes require updating conventions.md
• No duplicate concepts — if overlap exists, merge
• Cross-linking: modeling notes link to foundations; application notes link to modeling and engineering; reference implementations link to their conceptual counterpart
• Knowledge flows: Reading → Logs → Projects → Core layers (stable)
• Tags: cross-cutting metadata only; never restate the folder path (see §12 of conventions)

---

Getting Started

• Start at 00_meta/dashboard.md for active work
• Browse any layer’s index.md for entry points into that layer
• See 00_meta/glossary.md for term definitions
• Consult 00_meta/conventions.md before adding new notes

Link to original