GOATflow ingests raw operational intelligence — emails, documents, notes, voice memos — and returns prioritized, structured action items called Tracks. Built on Python & Streamlit, powered by Google Gemini, hosted on Streamlit Community Cloud.
Most productivity tools are passive record-keepers. GOATflow is active — it ingests raw operational intelligence (emails, documents, voice memos, images, pasted notes) and returns prioritized, structured action items called Tracks. Drop the intel into the Track Sieve. The Churn Engine does the rest.
The AI backbone — Google Gemini 1.5 Flash via the google-genai SDK — processes each submission and extracts Tracks with a title, description, priority tier (Micro / Standard / High-Leverage / GOAT), optional time estimate, and Summit Call flag for urgent items. It thinks like an operator’s chief of staff: not just extracting tasks but evaluating relative urgency and business impact.
A gamification layer — dual-currency economy, 7-level ascension system, real-time stat dashboard, and browser-native Goatifications — drives consistent follow-through. Built on Python 3 + Streamlit, hosted on Streamlit Community Cloud at zero monthly cost. Database on Supabase PostgreSQL. Invite-only access. This is a live production application with real auth and persistent data — not a demo.
Every screen below is from the live production prototype at goatflow.app — not mockups, not wireframes.
GOATflow runs a full dual-currency reward system built on behavioral reinforcement. Every Track completed earns Hay. At 500 Hay, the system converts automatically to Fresh Cheese — the premium currency tracked as CCR (Cheese Churn Rate). The Pasture Gauge at the bottom of the dashboard shows real-time progress toward the next conversion.
| Stat Card | What It Tracks |
|---|---|
| Active Tracks | Open items currently in the operational log |
| Summit Calls | Tracks flagged as urgent or critical |
| Completed | All-time Track completions |
| Hay | Current balance + progress to next Fresh Cheese conversion |
| Fresh Cheese | Total CCR (Cheese Churn Rate) earned |
| Active Horns | Count of live recurring prompt templates |
| Clip Rate | 7-day completion vs. generation ratio — color-coded green (≥75%), amber (≥50%), red (<50%) |
| GAIT | Consecutive daily completion streak |
| Stakes % | Today’s resolved Tracks vs. logged Tracks |
| Precision | % of timed Tracks completed within their estimate. Unlocks after 5 timed completions. |
st.markdown(unsafe_allow_html=True). JavaScript behavior is injected via streamlit-components at zero height.google-genai SDK. No AI is called at page load — only on explicit user submission. The prompt is tuned to evaluate urgency and business impact, not just extract text. Free tier: 15 requests/min, 1M tokens/day.ensure_schema(). Per-user data isolation across all tables: users, player, operational_log, signals, directives, horns, invites.UPDATE ... RETURNING prevents race conditions. Passwords hashed server-side. No third-party auth providers.streamlit run app.py. Secrets managed via Streamlit's secrets panel: DATABASE_URL, PROD_DATABASE_URL, GEMINI_API_KEY, SESSION_SECRET, ADMIN_USERNAME.
GOATflow is a subsidiary product of the WorkGOAT Ecosystem (workgoat.vip) — the enterprise-facing extension designed to bring the same AI operational intelligence and gamified accountability to team and organizational contexts.
Where GOATflow handles individual operational load, WorkGOAT scales the architecture to multi-user environments — enabling managers to route work intelligently across teams, track collective throughput, and build the kind of accountability culture that most project management tools fail to create.
GOATflow was conceived, architected, and shipped alongside a demanding operational career. It is one of five AI systems in this portfolio — each one built to solve a real problem, not to demonstrate technical curiosity.