A toolkit for managing GIS projects as code, built for working with an LLM. Treats a GIS project as build output: Python is the source of truth, and the same project spec renders to QGIS project files, ArcGIS Pro layer files, ArcGIS Online hosted layers, or static PNG maps. Describe a map change in plain English and let an LLM implement it; the project is reproducible from a fresh checkout and fully diffable in git.

Habitat suitability analysis for goat grazing at San José's Alum Rock Park. Combines a USGS elevation DEM, Santa Clara County vegetation data, and OpenStreetMap data to compute slope, classify vegetation by lifeform, and run a weighted overlay to identify grazeable patches. Ranks candidate staging areas by distance-weighted access to high-priority terrain. Built with the alidade toolkit. Also published as an ArcGIS story map.

An ArcGIS story map created for the GIST-11: Intro to Mapping & Spatial Reasoning course at Foothill College. A pair of maps with a slider compares pre-European and modern habitat extents. iNaturalist observations of key species map illustrate persistence of habitats, and Santa Clara County Parks trails enable public access. Switch to static PDF version when story map link expires.

Interactive and print maps for habitat restoration work areas in San José's Alum Rock Park, with context information sources from a Google Sheet.

Final project for SJSU GEOG 172 - Cartography and Data Visualization: a map comparing mean commute time in the San Francisco - San José urban area in 2000 and 2019. Built with R from census data and visualized with ggplot2.

Export Strava activities, convert to GeoJSON, and display on a Map Tiler map. Add layers for Bay Area Ridge and Bay Shoreline trails. Display picnic tables from Open Street Map.

A deep dive into working with GPS track data from Strava GPX files. Based on https://github.com/kielni/activity-map.

A history of geodata, from Hipparchus' spherical earth to John Harrison's marine chronometer to GeoJSON.

Draw markers on a Google map with API event data. Search for events with a Places search box. Show event data in popup on click of marker.

Building a map from a spreadsheet, with the Mapbox and Google Sheets APIs.

The Guadalupe River Trail is an amazing panorama of the city of San Jose.

Draw a map geo data (from TopoJSON URL, TopJoSON document on Google Drive, or GeoJSON document on Google Drive) plus fill color data (from Google Sheet)

A presentation covering passwords, passkeys, two-factor authentication, identity theft, data brokers, device hygiene, credit protection, phishing, and social media risks. Plain HTML and CSS, built with Reveal.js.

I set up an AWS Lambda function to that uses a PyToch model to detect when an aurora is visible on the AuroraMAX live feed, then send an alert so I can go outside and look. A web view displays images where a strong aurora was detected, along with a timeline for the past 24 hours.

I tried to make sense of wildly inaccurate solar + battery quote, then created my own Python model using pandas and lots of matplotlib charts. I modeled hourly flows from the grid and solar sources, to and from a battery, and exports to grid. I appplied complex PG&E credits, reviewed variations of arbitrage strategies, and illustrated how these affect cost and sourcing patterns.

A minimal AWS-hosted setup for saving articles, designed as a DIY alternative to Pocket.

Cut through the noise, save time, focus on the fun parts; that's the dream. AI agents to set filters and collect additional information from booking.com and Airbnb sites. The browser-use library seems promising, but it doesn't really work.

Brown bag presentation on natural language-driven browser automation, based on work in https://github.com/kielni/travel_agent.

Create summary sheets for college research with OpenAI API.

Python scripts to organize photos, including collecting photos from Apple Photos, extracting metadata, and archiving on S3. See also LivePhotos for the web, on transforming LivePhotos so they can be embedded on a page and played by any browser.

Planning to view the 2024 solar eclipse: why and where to go.

Get IPEDS college data into a human-friendly modern SQL database. Integrated Postsecondary Education Data System (IPEDS) is a system of interrelated surveys conducted annually by the National Center for Education Statistics.

Alviso the (mostly) forgotten port of San José.

Get car count data from SNAPS vehicle detection API and send to a Google Sheet.

Fairbanks cloud cover forecast, for aurora viewing planning.

A history of smart home automation: visions from the past and reality of today.

Google's AppSheet platform: No code / low code (was: wizards) + Sheets (was: Access) = app magic!

The story of Buzz Holstrom's epic solo journey 1,100 miles down the Colorado River through the Grand Canyon. "I find I have already had my reward -- in the doing of the thing."

Running production-scale tests with AWS Batch showed that merging data with SQLAlchemy reduced our Storage IO costs and accumulation of dead tuples.

Migrating from Postgres to Aurora was easy. Storage IO usage nearly doubled our costs. We reduced them by resizing instances and repacking tables.

Description of how I organize and share my photos, with metadata, codecs, Python script, and S3.

A guide to the Firebase realtime database, including real-time data synchronization, security rules, and data structures.

Randomly select 10 dinner ideas from Pocket saves tagged with "dinner." Solve the "what's for dinner?" question with saved recipes.

Alexa skill for Hillbrook School calendar.

A Chrome extension to set read date and rate a book on Goodreads from Kindle Cloud Reader.

A deep dive into Python logging, including handlers, formatters, and configuration. Also covers JSON, Docker, and celery logging.

Packrafting equipment and adventures in Silicon Valley.

A JavaScript application to browse and filter my Steam games library. Display logo, description, and tags, and hide or star games.

Generate searchable documentation for a tree of Starbound assets. Reads object and status effect files, generates documentation files, and runs a web server for browsing the content.

Visiting onsens in Japan.

Python worker bees for collecting and processing data. Aggregate RSS feeds into daily page with AWS Lambda and S3.

History of the Stanley Cup, the oldest existing trophy to be awarded to a professional sports franchise.

Silicon Valley has seasons. They're not the same as yours, and you need to get outside the city to experience them.

A simple Vue.js application maintain a list of TV shows in chronological order from last watched to most recent, backed by a Firebase realtime database.

Display per-store shopping lists, with realtime updates via Firebase.

In the olden days, software used to come in a box. Now, we can make sure our users always have fresh baked software. Ember apps in the cloud help.

The page object pattern makes writing acceptance tests for an Ember select2 component easier and faster.

A tale of robots, gold stars, and code coverage.

In a previous blog, I described how we deploy Ember apps to S3 buckets, and promised a follow up article about how we use a staging bucket to preview changes before pushing to production. In this article, I'll describe how we added shared development and staging buckets to our Ember deployment setup.

How serving Ember apps from S3 and Cloudfront simplified dev environments, sped up builds and deploys, and made our production infrastructure smaller and more scalable Gridium's Tikkit application has three separate front-end Ember apps that all talk to a common API.

Gridium's work order management system runs on a microservices architecture, with Docker containers managed by Kubernetes. We've divided our application into small, reusable components, each of which runs in its own Docker container.

Creating a prototype for a new product idea with AWS building blocks

Ignite talk about how and why to buy local grass-fed beef, in quarter-cow increments.