13h • General discussion
AI Developer Accelerator — Coaching Call - July 28th
AI Developer Accelerator — Coaching Call - July 28
Meeting Purpose
Key Takeaways
- EMSO Pivots Sales Strategy: To survive a cash crunch, EMSO is shifting from slow "whale" (fire department) deals to faster "tuna" (private ambulance) contracts, which convert in months vs. years.
- Patrick Automates Enterprise Support: A self-improving support skill for 2,500 Claude users is being built. It uses a feedback loop where support staff refine answers in Co-Work before they are pushed to a user-facing skill, ensuring accuracy and reducing support load.
- Scott's Security Tool GTM: The team advised Scott to package his advanced security review tool as a free plugin for viral distribution, using a freemium model to gather feedback and define a paid enterprise offering.
- Ryan's US Real Estate Opportunity: Ryan can enter the US market by productizing his AI video pipeline. He will use a simple "picture book" guide to enable non-photographers to capture high-quality assets, creating a scalable, high-margin service.
Topics
EMSO: Sales Pivot for Survival
- Problem: EMSO, an AI narrative tool for EMS/fire, faces a cash crunch. Its initial ICP (fire departments) has a 1.5-year sales cycle, which is too slow for its burn rate.
- Solution: Pivoting to a "tuna" strategy focused on private ambulance companies.
- Outreach: Using Loom videos and Instantly for asynchronous, scalable contact.
- ICP Shift: Focusing on private ambulance companies while using Instantly to manage "minnow" (small agency) leads.
- Internal Champions: Providing prospects with one-page presentations to pitch the value (e.g., "make you millions") to city councils.
Patrick: Scaling AI Support at Enterprise
- Problem: Scaling support for a Claude rollout from 250 to 2,500 users by fall.
- Solution: Building a self-improving support skill to automate Level 1 support.
- Process: Support staff refine answers in Co-Work → answers are added to a knowledge base → stable answers are pushed to a user-facing skill.
- Future: Integrate with ServiceNow for a RAG-based knowledge source.
Scott: Advanced AI Security Review Tool
- Process: Scans local diffs (secrets redacted) using three parallel engines: a custom agent suite, Anthropic's native review, and OpenAI's Codex CLI.
- Learning Loop: A dashboard tracks findings. Marking a finding "false positive" adds reasoning to future prompts, preventing repetition.
- Distribution: Package as a free plugin for viral marketplaces (e.g., Claude, Cursor).
- Monetization: Use a freemium model (e.g., free for 5 projects) to gather feedback and define a paid enterprise offering with custom governance rules.
Ty: Automated Dev Workflow & New App
- Process: An intern reports bugs via WhatsApp → an agent classifies the request → a Cloud Code session fixes the bug → the fix is deployed.
- Guardrails: A feature intent document prevents the agent from making out-of-scope changes.
- B2B Focus: Prioritize a B2B model (affiliate fees from courses) over B2C subscriptions.
- User Acquisition: Use ChatGPT ads to build a user base (e.g., 10k users for ~$3-4k) before approaching courses.
- Sales: Use Instantly for cold outreach to courses, focusing on one state at a time.
Morgan: Launching New Products
- GTM Strategy: Offer free "founder school" trials to build case studies, which are more valuable than early revenue.
- Sales: Use Loom videos for outreach, focusing on the outcome (e.g., "save 40% of dismissal time") to secure calls.
- Status: Map tool for creating plot layouts is complete.
- Next: Apply the same Loom-based sales strategy as for Class2Curb.
Ryan: Real Estate AI Video Opportunity
- Problem: Ryan's UK market has low pricing ceilings for video (£200-600), making a scalable AI service difficult.
- Opportunity: A high-margin US market exists for AI-generated real estate videos.
- Service: Generate promotional videos from existing high-res photos using an AI pipeline (Higgsfield CLI, Seedance).
- Execution: Create a simple "picture book" guide for non-photographers to capture the necessary assets.
- Pilot: Ryan will send example videos to Brandon, who will share them with his real estate agent sister to test the US market.
Next Steps
- Implement the Codex plugin in Claude Code for automated cross-model reviews.
- Upgrade sales presentations to interactive HTML using Claude Design and Google Omni.
- Email Brandon example real estate videos to explore the US market opportunity.
- Create the "picture book" guide for capturing real estate assets.
- Package the security review tool as a free plugin for viral distribution.
- Plan a multi-tenant architecture to support a freemium model.
- Launch outreach campaigns for Class2Curb and Heritage Plot using Loom videos.
- Secure "founder school" trials to build case studies.
Action Items
- Research city council/board outreach tactics for fire/EMS - WATCH (5 secs)
- Research EMS grants; share findings w/ Brandon - WATCH (5 secs)
- Record and share Agent Tax demo videos w/ Brandon - WATCH (5 secs)
- Send Claude Code status line link to Brandon - WATCH (5 secs)
- Create SOP for AI photo booth install/uninstall - WATCH (5 secs)
- Productionalize AI photo booth backend for scale/availability - WATCH (5 secs)
- Record 3–5 Class2Curb Loom outreach videos daily - WATCH (5 secs)
- Email Brandon real estate video examples + photo checklist - WATCH (5 secs)
- Convert outreach decks to interactive HTML slides via Claude Design - WATCH (5 secs)
0
0 comments
AI Developer Accelerator — Coaching Call - July 28th
skool.com/ai-developer-accelerator
Master AI & software development to build apps and unlock new income streams. Transform ideas into profits. 💡➕🤖➕👨💻🟰💰
Leaderboard (30-day)
1
+6
2
+6
3
+4
4
+3
5
+2
Powered by