Day in Our Data · Saturday, October 3, 2026, 11 a.m. to 3 p.m.
Two ways to take part. Pick a track, get set up in about 15 minutes, and spend the rest of the day on Oak Park's data. AI tools are welcome, and so is a project that uses none.
None of this is required. Each item saves you ten minutes at your table.
Both tracks end in the same place: an entry on this site. They differ in how much you set up. If you are not sure, take Track A.
Track A · Zero setup · Recommended
A ready-made coding environment built for this event. It runs in your browser, the starter projects are already loaded, and you cannot break anything. No GitHub account needed: download your project as a zip and attach it to your entry.
Track B · Your own tools
Claude Code in your terminal or editor, Claude Code on the web, Python or Jupyter, or plain Git with no AI at all. Your work lives in a GitHub repository you link from your entry.
Go to https://day-in-our-data.fly.dev/ and create an account with any email address you can check.
Open the Agent tab and paste your key first. If you start a session before entering a key, the agent has nothing to run on. Your table host hands you the GLM key at check-in.
For GLM, paste the event key. For Claude, paste your own Claude token and pick Opus as the model.
Type a plain-English request in the Agent tab. Try: List the starter projects in this workspace and tell me which one a beginner could finish in two hours.
Prefer typing commands? In the Terminal tab, run opencode for the GLM agent, or claude for Claude Code after you have entered your Claude token.
opencode # GLM agent in the terminal
claude # Claude Code in the terminalBefore you leave, and before you enter your project, download your project from the workspace as a zip. That zip is your copy: it is what you attach to your entry, and it outlasts the workspace. Can't find the download option? Ask a table host.
If you built a web page, the workspace can show you a preview link. Open it in a private browser window first: if it loads without signing in, paste it into your entry as your demo link.
A preview link only works while your workspace is running. Voters open links after building ends, so the zip is what counts. A screenshot in the zip helps too.
Track B assumes a free GitHub account: your repository is how teammates share work and how you link it from your entry. Pick whichever tool below suits you.
Most control
Runs on your own files. Install steps are below. Works inside VS Code, Cursor, and JetBrains IDEs too: see where to code.
No install
Use claude.ai/code in the browser, or the Code tab of the Claude desktop app. Needs a Pro, Max, Team, Enterprise, or Console account; the event credits apply to a Console account. Connect GitHub when asked and pick your team's repository.
Your usual stack
Clone or download the event repository for the starter projects and cached data, and work however you normally do. Commit to your team repository as you go.
Claude Code runs on macOS 13+, Windows 10 (1809) or later, and most Linux distributions, with 4 GB of memory or more. You do not need administrator rights.
Mac: press Cmd+Space, type Terminal, press Enter. Windows: press the Windows key, type PowerShell, press Enter. Linux: you know where it is.
Copy the one line for your system, paste it into the terminal, press Enter.
# macOS, Linux, or WSL
curl -fsSL https://claude.ai/install.sh | bash
# Windows PowerShell
irm https://claude.ai/install.ps1 | iex
# Or with a package manager
brew install --cask claude-code # macOS Homebrew
winget install Anthropic.ClaudeCode # WindowsWindows: installing Git for Windows is optional but lets Claude Code use Git Bash. If you see 'irm' is not recognized, you are in CMD, not PowerShell.
Close the terminal, open a new one, and run the version check. If it says claude is not found, open a fresh terminal once more, then ask a table host.
claude --version # prints a version number
claude doctor # checks your install and settingsDownload the event repository, move into it, and start Claude Code. The first run opens your browser to sign in.
git clone https://github.com/oak-park-cisc/Oak_Park_Day_in_our_Data
cd Oak_Park_Day_in_our_Data
claudeNo git? Download the repository as a ZIP from GitHub (green Code button, then Download ZIP), unzip it, and cd into the folder.
Install the Claude Code extension for VS Code (it also works in Cursor), or the plugin for JetBrains IDEs. See where to code for the trade-offs.
Same steps on both tracks, whichever agent you use. Type these in plain English. You do not need to know any code.
Run /init. Claude reads the files and writes a CLAUDE.md: notes it will reread at the start of every session.
Explain what is in this project like I have never coded before. What data is here and what could I make with it?
Press Shift+Tab until the mode shows plan. Claude now researches and proposes a plan without changing any files. Say Plan a simple map for starter project 11 (the urban forest). Read the plan, push back, then switch out of plan mode and say Go.
Show me how to open what you just built. Claude will give you a command, a file to open, or a link.
Put every step that changes the data in a script or notebook, not just in this chat, and tell me which file does what. A result nobody can rerun is hard to trust, and the judges check for it.
Start this by 2:00. Your README is how the AI judging panel understands your project: it reads your entry and your README, not your code or your app. Use these five headings, one short paragraph each.
| Heading | What to write |
|---|---|
| The question | What resident question does this answer, and who asks it? |
| The data | Which datasets you used, where they came from, and what they do not cover. |
| How we got the numbers | Each step from the raw file to your result: filtered, joined, counted, charted. Say which script or spreadsheet does it, and where AI helped. |
| How to use it | How to open it or read it, in plain words. |
| Status and next steps | What works, what does not yet, and what someone should do on Monday. |
Write README.md for this project using exactly these five headings:
The question, The data, How we got the numbers, How to use it,
Status and next steps. Only describe what is really in these files.
List every data file we used and every step that changes the numbers.
Keep it under one page.README.md. Replace it; the placeholder counts as no README. It then travels inside your zip.README.md at the top level of your repository.Every team enters on this site, once, using the entry form. Nothing else to fill in anywhere.
| Field | Track A (workspace) | Track B (your own tools) |
|---|---|---|
| Written answers | Required | Required |
| How you got from the raw data to the result | Required: a few plain sentences | Required: a few plain sentences |
| Repository link | Leave blank | Your GitHub repository |
| Live demo link | Optional: your preview link, if it opens signed out | Optional: anywhere public |
| Upload | Your project zip (up to 8 MB) | Optional: a chart, CSV, PDF, or zip |