AI Developer Accelerator
Log In
Community
Classroom
Calendar
Members
Leaderboards
About
Log In
5
Patrick Chouinard
11d •
General discussion
AI Developer Accelerator — Coaching Call - August 11th
AI Developer Accelerator — Coaching Call - August 11
VIEW RECORDING - 96 mins (No highlights)
Meeting Purpose
To share progress, discuss AI agent architectures, and strategize on enterprise adoption.
Key Takeaways
Chief of Staff (CoS) Architecture: Multiple members are adopting a CoS model, shifting from reactive assistant use to a proactive coordinator that manages tasks, prioritizes work, and initiates agent sessions, freeing up human time for high-level decisions.
Enterprise Adoption Strategy: Overcoming developer skepticism requires proving ROI with direct performance comparisons (e.g., agentic vs. traditional dev delivery) and enforcing governance via mandatory skills and plugins, not just rules.
Convex.dev
for Real-time Apps:
Convex.dev
is being considered for a carpool app to replace Supabase. Its native real-time sync simplifies architecture, but indexing queries is critical to manage costs and prevent bill shock.
New Business Models: AI enables new models like Juan's low-cost ($1.5k) AI photo booth, which may pivot from VC funding to a distribution partnership with event venues.
Topics
AI Agent Architectures
Problem: Managing multiple AI assistants becomes a bottleneck, consuming more time coordinating agents than doing the work itself.
Solution → Chief of Staff (CoS) Model:
Coordinator (CoS Agent): Manages tasks, priorities, and agent coordination.
Executor (Code Agent): Performs the actual work (e.g., coding, writing).
Human Role: Becomes a high-level decision-maker, approving plans and providing direction.
Implementations:
Patrick: Uses Cowork (CoS) and Claude Code (Executor) communicating via a local Git repo. Cowork monitors Teams, calendar, email, and Jira to provide daily briefings and prompt Claude Code sessions.
Paul: Uses a similar CoS model, requiring human approval for key decisions to prevent overwhelming noise from multiple agent activities.
Ty: Also shifting to a CoS model, focusing on seeing only problems that require human intervention.
Ryan: Setting up "Hermes" as a "Digital Personal Assistant" to handle administrative tasks (e.g., scheduling, calls), freeing up time for creative work.
Enterprise Adoption & ROI
Challenge: Overcoming developer skepticism and proving ROI for AI tools in an enterprise setting.
Strategies:
Direct Performance Comparison: Prove ROI by having an agentic-aided developer and a traditional developer swap projects. The agentic developer's 3x faster delivery provides undeniable proof of performance.
Governance via Skills: Enforce enterprise standards and tech stacks through mandatory skills and plugins, not just rules.
Example: An onboarding skill can automatically structure a user's workspace and enforce directory limits.
"Living Specifications": Reframe prototypes as "living specifications" by embedding functional spec data as comments in the HTML output. This allows a code agent to use the output as a direct blueprint for building a real application.
ROI Measurement: Focus on the value of the end deliverable ("solution per token"), not proxy metrics like lines of code or token consumption, which are easily gamed.
Project Updates & Technical Deep Dives
Ty:
Family History Project: Built a web app from WhatsApp group chats to visualize family trees, locations, and birthdays. The app updates daily and is editable by family members.
AI Golf Coach: Analyzes Garmin watch data and shot outcomes to provide swing coaching. Identified patterns like poor shots after waiting >5 minutes or making fun of others.
mdcatc:
Carpool App Rewrite: Considering a rewrite to
Convex.dev
from Supabase for its native real-time sync.
Rationale: To support 300+ concurrent clients (e.g., at school dismissal) without managing complex socket subscriptions.
Bastian's Advice: Use Convex's WorkOS integration for multi-tenancy and index queries to manage costs.
Book Automation Workflow: An agent process converts raw manuscripts into four outputs: paperback, hardback, e-book, and audiobook.
Paul:
Driver Verification Project: Using the Veriff API for biometric validation (ID + live face check) to solve a client's problem of drivers using fake identities.
Ryan:
Video Generation: Creating a 1-minute animated video (10-12 scenes) using a modified repo, ElevenLabs for voice-over, and Higgsfield CLI.
Juan:
AI Photo Booth: Refining the setup for upcoming events.
Hardware Cost: ~$1,500 (mini PCs, mirror booth).
Funding Strategy: Exploring a pivot from VC to a distribution partnership with event venues, offering them a revenue cut.
Next Steps
mdcatc: Decide on the
Convex.dev
rewrite for the Carpool app.
Juan: Connect with Paul to discuss funding strategies and with Ty to explore distribution models for the photo booth.
Ryan: Share the completed animated video with the group next week.
Scott: Demonstrate the new Cursor/VS Code replacement tool at the next meeting.
Action Items
Draft Claude style templates for Chief of Staff + Claude Code -
WATCH (5 secs)
Decide this week: rewrite Carpool app to
Convex.dev
-
WATCH (5 secs)
Schedule 1:1 w/ Paul re: funding/distribution -
WATCH (5 secs)
DM Ty re: mirror booth distribution/licensing -
WATCH (5 secs)
Share 1-min animated video w/ group next week -
WATCH (5 secs)
DM Patrick re: Hermes setup/memory tips -
WATCH (5 secs)
Like
1
0 comments
5
AI Developer Accelerator — Coaching Call - August 11th
AI Developer Accelerator
skool.com/ai-developer-accelerator
Master AI & software development to build apps and unlock new income streams. Transform ideas into profits. 💡➕🤖➕👨💻🟰💰
11.4k
Members
54
Online
8
Admins
JOIN GROUP
Leaderboard (30-day)
1
Tom Welsh
+3
2
Ali BabaStudios
+3
3
Akash Narayan
+2
4
A. Bhardwaj
+1
5
Parvez Ahmed
+1
See all leaderboards
Powered by