LLLogan LamCloud / Data / AI Systems

Projects

Work that shows the system behind the interface

These projects are intentionally framed around the problem model, technical boundaries, and lessons learned. Some are implementation work, while others are structured learning or architecture concepts.

Full-stack SystemIn Progress

Leaf Creme

A full-stack commerce-style system focused on backend logic, database modeling, and frontend-backend consistency.

FastAPIPostgreSQLSQLAlchemyReactTypeScriptTailwindMUI

Focus

  • API design
  • Database modeling
  • Business logic
  • Frontend-backend integration

Learning

The hardest part of full-stack development is keeping the mental model, database model, API contract, and interface aligned.

Cloud LearningIn Progress

AWS Learning Journey

A collection of notes, labs, and reflections from AWS learning and FCAJ AWS Bootcamp.

AWSIAMCloud FundamentalsSecurityArchitecture Notes

Focus

  • Cloud fundamentals
  • IAM and security basics
  • Deployment mindset
  • Scalable infrastructure thinking

Learning

Cloud is not just renting servers. It changes how we think about reliability, scalability, security, and cost.

AI-native WorkflowExploring

AI Workflow Experiments

Experiments using AI tools to design, debug, and document software systems with better context and constraints.

ChatGPTCodexPrompt EngineeringContext Engineering

Focus

  • AI-assisted development
  • Prompt design
  • Context engineering
  • Human-in-the-loop development

Learning

The quality of AI output depends less on how much context we give, and more on how clearly we define goals, constraints, and decision boundaries.

System Design ConceptConcept

Startup Credit Scoring Multi-Agent Concept

A system design reflection inspired by enterprise-grade multi-agent architecture for startup credit scoring.

Multi-Agent SystemsBusiness RulesComplianceRisk Analysis

Focus

  • Single-agent vs multi-agent decision design
  • Virtual credit committee model
  • Guardrails and compliance
  • Risk analysis

Learning

Some business problems are too complex for a single model or single workflow. Multi-agent systems become useful when the problem requires specialized judgment, separation of concerns, and controlled decision-making.