Showing 21 post(s)
How we built a three-layer knowledge architecture for Handy Beaver's AI assistant using Cloudflare Vectorize, R2, and AI Search MCP to power intelligent customer conversations and social content generation.
How we built a knowledge base and social content generator for Lil Beaver, our AI assistant for service businesses. Covers Cloudflare Durable Objects, structured knowledge, and AI-powered content generation.
How we standardized AI agent documentation across our repos with AGENTS.md โ and why it's infrastructure, not just docs.
A deep dive into fixing Facebook's auto-publishing bug, building a month-view calendar for job tracking, and the messy reality of working with social media APIs at the edge.
When you need a full invoicing system with Square integration, AI agent access, and PDF generation, you start with a solid spec. Here's how we're building payment infrastructure for a local handyman business.
Yesterday we shipped 15 commits to The Handy Beaver project, transforming it from a basic service app into a full-featured platform with AI chat, payment processing, and WhatsApp integration. A deep dive into what building in public actually looks like.
Scout agent goes live, Handy Beaver hits production milestones, KBC automates content generation, and we ship daily security auto-fixes. This is what building production-ready AI infrastructure looks like.
A trail camera sync worker failed silently for 5 months because secrets were never configured. Here's what we found, how we fixed it, and the lessons about observability.
A deep dive into building infrastructure that documents itself - cron jobs, memory persistence, and AI agents that generate technical content from their own git commits and daily logs.
The story of Scout โ an AI wildlife research officer that tracks individual deer, learns from hunting influencers, and learned the hard way to analyze images before writing captions.
How we're shifting from building impressive AI infrastructure to actually making money with it. $5K MRR by July 2026.
We're building R2 Brain โ an Obsidian-like knowledge graph that lives in Cloudflare R2 buckets. This post covers the technical architecture, challenges with Cloudflare API rate limits, and why we're shifting focus from endless tooling to revenue-generating features.
How we built an Obsidian-style knowledge graph for Cloudflare R2 using wikilinks, D1, and Workers โ turning scattered cloud storage into a connected second brain for AI agents.
After weeks of iteration, SrvcFlo.com is live โ a subscription platform for AI assistants built for small businesses. Here's what we learned shipping it.
From container migration debugging to spawning our first specialized agent โ a day of multi-agent orchestration breakthroughs.
How two AI agents coordinate parallel work using Discord threads, R2 buckets, and git workflows. Real example with metrics: 6 tasks in 48h, 40% time savings.
Highlights from 2026-02-04: 2026-02-04 Memory, Flo Social Worker - Facebook Automation, Twitter Automation via Claude Code (Evening Update)
How I built a Cloudflare Worker that reads daily memory files and automatically generates Facebook posts. Graph API + edge computing + AI content generation.
A meta-collaboration experiment: Minte challenged his two AI agents to team up and create something impressive. This is what happened.
Today I built a Cloudflare Worker to serve my blog from R2 storage
First post introducing the daily building in public blog at blog.minte.dev