Research Intelligence Systems

Clean-room public-data research workflows for content gaps, market monitoring, narrative scanning, and opportunity discovery.

This repository is a portfolio-safe demo for recruiter and collaborator review. It shows how I structure research automation systems while keeping private strategies, accounts, credentials, and live execution paths out of public code.

What this demonstrates

• transparent weighted scoring over public-style signals
• synthetic fixtures for content-gap, narrative-scan, and market-monitor examples
• redaction of common sensitive strings before reporting
• approval-gated queues that stop at draft_report_only
• markdown report generation
• tests and GitHub Actions CI

Safety boundaries

This repo is intentionally dry-run only. It does not include:

• credentials, cookies, API keys, tokens, wallets, or account IDs
• private source lists, strategy thresholds, logs, or screenshots
• code that trades, submits forms, publishes posts, spends money, or mutates accounts
• production scraping, private endpoints, or proprietary edge logic

See docs/safety-boundaries.md.

Quick start

python3 -m venv .venv
. .venv/bin/activate
pip install pytest
PYTHONPATH=src pytest -q
PYTHONPATH=src python -m research_intel.demo examples/content_gap_input.json

Generate all sample reports:

mkdir -p reports/generated
for f in examples/*_input.json; do
  base=$(basename "$f" _input.json)
  PYTHONPATH=src python -m research_intel.demo "$f" > "reports/generated/${base}_report.md"
done

Author

Thaddeus Arndt / Preballin

• Portfolio: https://www.preballin.com/portfolio/
• GitHub: https://github.com/thaddeusarndt