I Turned My Obsidian Vault Into Claude's Brain (And It Took Less Than a Day)
My Obsidian vault and Claude Code were living completely separate lives. So I connected them: 20 slash commands, auto session summaries, git-backed notes, live MCP. The whole thing runs on plain text and took an afternoon.
Read More → Building a RAG System for Large Text Files: Lessons from the Trenches
I tried to build a search tool that only returns exact quotes from uploaded text files. What followed was a crash course in free-tier limits, dependency hell, and knowing when to stop.
Read More → Why I Moved from SQL to Databricks (And Never Looked Back)
A practical look at what made me switch from traditional SQL workflows to Databricks: the good, the painful, and the game-changing.
Read More → I Run Three AI Agents as My Dev Team (And Here's Where It Breaks)
Claude Code handles reactive work. Hermes runs autonomous missions with persistent memory. Gemini CLI does cheap parallel reads. The architecture is coherent. Here's where it's probably over-engineered.
Read More → Building a 160-Chapter Knowledge Base with Free LLMs and OCR'd Reference Books
How I turned 3,000+ pages of scanned reference books into 160 structured tutorial chapters for under $1. The solution: a multi-model editorial pipeline, local OCR extraction, and Claude Code as the orchestrator.
Read More → Dashboard Design That People Actually Use
Lessons from building dashboards that stakeholders actually open more than once. Hint: less is more.
Read More → The Data Pipeline Nobody Talks About: ETL in the Real World
ETL sounds simple in theory. In practice, it's where data engineers earn their battle scars. Here's what I've learned.
Read More → How AI Tools Changed My Analytics Workflow
From writing SQL to letting LLMs help me debug pipelines: how AI assistants have become part of my daily toolkit.
Read More → Getting Started with SLAM for Autonomous Navigation
A beginner-friendly guide to Simultaneous Localization and Mapping: the algorithms, the math, and how to actually build one.
Read More → Fine-tuning LLMs for Domain-Specific Tasks
Practical lessons from fine-tuning language models: when it works, when it doesn't, and what I'd do differently.
Read More → Building a Real-Time Sensor Fusion Pipeline
How I combined LIDAR, IMU, and camera data into a single coherent stream for a robotics project.
Read More → Why Every Engineer Should Learn Embedded Systems
Understanding hardware makes you a better software engineer. Here's why, and how to get started without buying a single board.
Read More →