Engineering
Insights
Technical deep dives into architecture, Agentic AI, and full-stack engineering.
RESULTS: 32 UNITS
Designing a Deterministic AI Workflow Runtime in Rust
Why I built the Agentic SDLC runtime in Rust, how I achieved deterministic execution with trace IDs and an LLM replay store, and what 183 passing tests taught me about AI reliability.
Connecting AI Agents to the Real World: APIs, Databases, and File Systems
A practical guide to tool and function calling for AI agents — from safe file system access to database queries to sandboxed execution, with a look at the emerging MCP protocol.
Building Production AI Agents with LangChain: Memory, Tools, and Reliability
How I built a real AI agent system for the Agentic SDLC project using LangChain, covering vector-based memory, tool integration, and reliability patterns that actually hold up in production.
REST vs gRPC: Choosing the Right API Pattern for Your System
A practical decision framework for choosing between REST and gRPC, with real code comparisons, versioning strategies, and honest trade-offs from production experience.
Clean Architecture with TypeScript and Node.js: A Practical Guide
A ground-up guide to implementing Clean Architecture in a TypeScript Node.js project — real folder structure, dependency inversion, repository pattern, and use cases that actually make sense.
Containerizing a Legacy Node.js App with Docker: Lessons Learned
Hard-won lessons from containerizing a legacy Node.js application — multi-stage builds, shrinking image size by 80%, handling legacy dependencies, and setting up docker-compose for local development.
Building Your First Cross-Platform Desktop App with Electron.js
Learn the basics of Electron.js and how I used it with Jitsi Meet to build the Naver Cloud Meeting desktop app for Windows, macOS, and Linux.
Authentication Patterns in Next.js: JWT, Sessions, and OAuth 2.0
A practical comparison of JWT vs session-based auth in Next.js, with real implementation patterns for OAuth 2.0 with NextAuth, secure cookie handling, and refresh token rotation.
Building a Production CI/CD Pipeline with GitHub Actions
A practical walkthrough of building a production-grade CI/CD pipeline with GitHub Actions — the real workflow I use to deploy my portfolio to AWS Lightsail, including self-hosted runners, secrets management, and branch protection.
Practical Prompt Engineering Patterns for Production LLM Systems
The prompt engineering techniques I actually use in production — chain-of-thought, few-shot, structured output, and how to handle hallucinations and balance cost against quality.
The Monolith to Microservices Journey: Strangler Fig Pattern in Practice
A practical walkthrough of migrating a monolithic Node.js application to microservices using the Strangler Fig pattern, with lessons learned from real production migrations.
Multi-Agent Orchestration with CrewAI: Roles, Tasks, and Delegation
Why single AI agents hit a wall on complex tasks, and how I used CrewAI to build a multi-agent system for the Agentic SDLC project with clear roles, task delegation, and shared memory.