Prizes:

1st Place: $5,000 Cash

2nd Place: $2,000 Cash

3rd Place: $1,000 Cash

The Goal:

Build services that analyze a user's own exported data and give them something they couldn't see before.

The Problem:

Think of all the data a person can already download today — emails, AI chat logs, social media archives, calendar history, financial transactions. What insights are buried in there? What patterns? What could you build that helps a real person understand themselves, their habits, their finances, or their relationships better?

Example Projects

• A tool that analyzes a user's ChatGPT conversation history to surface recurring anxieties, growth themes, and decision patterns
• A freelancer's "business brain" that ingests invoices, emails, and calendar data to auto-generate a weekly income + pipeline summary
• A service that combines social posts and transaction data to help a user understand their spending triggers and distraction patterns

Dataset

Issued Construction Permits (Austin) (2M+ rows of text)

<aside> <img src="/icons/light-bulb_orange.svg" alt="/icons/light-bulb_orange.svg" width="40px" />

About

This track benefits most from combining multiple data sources for a single person. The real magic happens at the intersections — what does someone's spending look like in the weeks after high-stress calendar periods? Do their social posts reflect what they're actually dealing with in their lifelog? Each persona's files cross-reference each other via refs fields, making these patterns discoverable. All data types a real person can export today (AI history, email, social, financial, calendar) are represented.

Any persona works well for this track. p01 (Jordan Lee) and p05 (Theo Nakamura) have particularly interesting financial + professional tension that's rich for analysis.

</aside>

<aside> <img src="/icons/code_orange.svg" alt="/icons/code_orange.svg" width="40px" />

Accessing the Datasets

→ Download from Google Drive

Quick load — combining sources for cross-analysis (Python):

python

`import json from collections import Counter

with open("persona_p05/transactions.jsonl") as f: transactions = [json.loads(line) for line in f]

with open("persona_p05/social_posts.jsonl") as f: social = [json.loads(line) for line in f]

with open("persona_p05/conversations.jsonl") as f: conversations = [json.loads(line) for line in f]

What does this person spend on most?

spend_tags = [tag for t in transactions for tag in t["tags"]] print(Counter(spend_tags).most_common(10))

What topics come up most in their AI sessions?

ai_tags = [tag for c in conversations for tag in c["tags"]] print(Counter(ai_tags).most_common(10))`

See QUICKSTART.md in the dataset folder for full examples.

</aside>

<aside> <img src="/icons/code_orange.svg" alt="/icons/code_orange.svg" width="40px" />

Using Your Own Real Data (optional)

• 💬 ChatGPT export — help.openai.com
• 🌐 Google Takeout — takeout.google.com — Gmail, Calendar, Drive, Location History
• 📱 Instagram / Facebook — Settings → Your Activity → Download Your Information (JSON format)
• 🍎 Apple Health — Health app → Profile → Export All Health Data

See how_to_export_your_own_data.md in the dataset folder for step-by-step instructions.

</aside>