Sharing with the community since 2007
My personal corner of the web — writing, open-source projects, agents, datasets, and talks from two decades of building data platforms. True to form, the whole site is data too: every list here is a parquet file.
Latest on LinkedIn
All →
LinkedIn ↗Jan 11, 2026Medallion Architecture & Star Schema: Old Concepts Powering Modern AI Pipelines
You have a flat file or a table. You need metrics. So you write some joins, add aggregations, and ship it -> Nothing wrong with that, It works. But as data grows and questions multiply, that approach starts to crack. To illustrate a more sustainable path, I'll use an airports dataset — 19,793 records, 27 columns mixing GPS coordinates, facility codes, and certification details. Typical messy sourc
LinkedIn ↗Jan 9, 2026Naming Conventions I Use Daily.
For a long time, I have written code, named files, and defined configuration and environment variables using a consistent style that felt natural. While I was familiar with terms like camelCase, PascalCase, and kebab-case, I often mixed up or forgot others. Here is my personal cheat sheet: Blog post URLs like my-awesome-post.md ? That's kebab-case. Environment variables like API_SECRET_KEY? That's
LinkedIn ↗Jan 2, 2026Part 2 - Loading CSV Data into Apache Iceberg Tables with PyIceberg and PyArrow
In Part 1 - Building a $0 Data Lakehouse with Apache Iceberg, Nessie, Neon, and Cloudflare R2 , we set up the infrastructure — a Nessie catalog on Fly.io, Neon PostgreSQL for metadata, and Cloudflare R2 for object storage. All on free tiers. Now comes the interesting part: actually loading data into Iceberg tables . In this article, I'll walk you through a Python-based approach that: Reads CSV fil
LinkedIn ↗Jan 2, 2026One-Line Local Web Server with Python
Sometimes you just need to quickly serve some HTML files locally. Maybe you're testing a static site, previewing an SVG, or checking how your CSS looks in a browser. You don't need to install anything. Python has you covered. The Command That's it. One line. What It Does Serves files from your current directory Runs on http://localhost:8888 (or any port you specify) Zero configuration - just run a
LinkedIn ↗Dec 31, 2025Looking Back on 25 Years in Software Development
This year marks 25 years since I started in software development. As I look back, I'm filled with gratitude for everyone who shaped this journey. To my mentors who saw potential when I couldn't, to colleagues who challenged me to think differently, and to the open source community that welcomed contributions from a curious developer — thank you. Every code review, every late-night debugging sessio
LinkedIn ↗Dec 31, 2025Part 1 - Building a $0 Data Lakehouse with Apache Iceberg, Nessie, Neon, and Cloudflare R2
Modern data lakehouses typically require expensive infrastructure—object storage, metadata catalogs, and databases. But what if you could build a production-ready Apache Iceberg lakehouse using only free-tier services ? In this article, I'll show you how to set up a complete data lakehouse stack for $0/month using: Nessie - Git-like catalog for Apache Iceberg ( Fly.io free tier) Neon - Serverless
LinkedIn ↗Feb 3, 2025Edge-Optimized GenAI Chat Web-UI for Ollama & Cloudflare AI models with Context Injection
It is a simple GenAI chat web interface built with Next.js that connects to both local Ollama and Cloudflare AI models. This application provides an intuitive interface for interacting with various AI models through a generic chat completions API, featuring customizable settings and real-time streaming responses. The default selected model is Technology Stack Frontend : Next.js 14, React 18 Stylin
LinkedIn ↗Jan 27, 2025Embedding LLM in SQL is simple!
Inspired by Snowflake’s AI-driven analytics and MotherDuck’s recent implementation of SQL + LLM ( Introducing the function), I explored LLM-powered Business Insights using DuckDB , embedding AI directly into SQL workflows for real-time data intelligence . Use Case: Privacy Engineering – Detecting and Masking PII The fully working code is available as a Jupyter Notebook , which can be accessed from
LinkedIn ↗Jan 6, 2025Data Pipeline Diagrams as Code
Introduction Creating diagrams for data pipelines, workflows, and architectures is a routine task for many of us. Whether it's for internal presentations, client discussions, or technical documentation, the process can often feel repetitive and time-consuming. I just launched a new open source tool/utility designed to simplify the process of creating data pipeline diagrams—a YAML-based tool that g
LinkedIn ↗Dec 31, 2024Privacy Engineering with LLMs: Safeguarding PII While Enhancing Text with Privacy Shield
Introduction In today’s data-driven environment, the use of Large Language Models (LLMs) for generative AI and text processing is not just important but increasingly common. With tools like ChatGPT, Google Gemini, Claude.ai , Perplexity and open-source AI models like Llama, Mistral.ai etc. people now rely on LLMs for tasks ranging from email proofreading to refining documents like legal contracts
Recent blog posts
All →Building a Generative Q&A Engine with Semantic Search and Embeddings
BlogApr 22, 2025Building a Generative Q&A Engine with Semantic Search and Embeddings
In the age of AI-powered tools and intelligent systems, the ability to search, analyze, and generate insights from documents has become paramount.
Edge-Optimized GenAI Chat Web-UI for Ollama & Cloudflare AI models with Context Injection
BlogFeb 3, 2025Edge-Optimized GenAI Chat Web-UI for Ollama & Cloudflare AI models with Context Injection
It is a simple GenAI chat web interface built with Next.js that connects to both local Ollama and Cloudflare AI models.
Embedding LLM in SQL is simple!
BlogJan 27, 2025Embedding LLM in SQL is simple!
Inspired by Snowflake’s AI-driven analytics and MotherDuck’s recent implementation of SQL + LLM (Introducing the prompt() function), I explored LLM-powered Business Insights using DuckDB, embedding AI directly into SQL workflows for real-time data intelligence.
Data Pipeline Diagrams as Code
BlogJan 6, 2025Data Pipeline Diagrams as Code
Creating diagrams for data pipelines, workflows, and architectures is a routine task for many of us. Whether it's for internal presentations, client discussions, or technical documentation, the process can often feel repetitive and time-consuming.
Privacy Engineering with LLMs: Safeguarding PII While Enhancing Text with Privacy Shield
BlogDec 31, 2024Privacy Engineering with LLMs: Safeguarding PII While Enhancing Text with Privacy Shield
In the age of AI-powered tools and intelligent systems, the ability to search, analyze, and generate insights from documents has become paramount.
Lightweight HTML to PDF with Puppeteer: Simple PDF Export for Edge Deployments
BlogDec 27, 2024Lightweight HTML to PDF with Puppeteer: Simple PDF Export for Edge Deployments
This microservice is designed to address simple PDF export needs, making it ideal for generating blog articles, profile cards, and other lightweight use cases.
Projects
All →AI Agents Monorepo
Monorepo of Vercel Eve agents (API test generator, diagram generator, GitHub PR digest, LinkedIn cover generator) sharing a common runtime kit, plus the AI SDET Workbench slide deck.
largeOSSAIAgentsTypeScriptLive demoClaimWise
Healthcare RCM synthetic data generator and data lakehouse demo monorepo — inspired by dbt's jaffle_shop.
largeOSSHealthcareData-EngineeringdbtTemplrGo
Lightweight developer platform with Lua serverless functions, SQL playgrounds, embedded BoltDB storage, and AI chat assistant. Config-driven, single-binary deployment.
largeOSSGoPlatformCMSLive demoDagster Data Pipelines
Collection of data engineering projects using Dagster to manage and orchestrate data pipelines.
largeOSSPythonDagsterData-Engineering