Build in public: my current Good AiDeas setup
I’m building Good AiDeas to help small local businesses get access to the kind of tech, systems, and process improvements that bigger companies usually get/can afford first. I’m still early but I wanted to share the actual setup I’m using right now. Hardware - MacBook Pro M5 - 64GB unified memory - Local Ollama experiments for lightweight background AI tasks - Looking at a stronger local AI workstation path later, likely NVIDIA or AMD-based, once the business justifies it Core AI tools - Codex for coding, research, documentation, debugging, and turning ideas into working pieces - Claude Code for engineering help, code review, refactoring, and deeper implementation work - MiniMax for creative/media experiments and content assets (cheaper than the other 2) - Local models/Ollama for smaller organizer-style workflows and internal learning loops Infrastructure - GoodAiDeas.com is live - VPS setup for hosting, deployments, mail/workflow experiments, and internal tools - Internal systems for lead tracking, email workflow, AI receptionist, and service-business playbooks - I keep sensitive things like IPs, credentials, and internal routes private for obvious reasons Current product direction Right now I’m focused on practical AI for small businesses, especially service businesses. Things like: - missed calls - after-hours intake - lead follow-up - owner daily summaries - quote/estimate follow-up - invoice and review workflows later The goal is not “AI everywhere.” The goal is to find the spots where business owners are losing time, money, or opportunities, then build systems that help. Social setup I’m also starting to build in public more across: - LinkedIn - Facebook - X.com - Skool/community posts - GoodAiDeas.com I want to show the process honestly: What I’m building, what works, what breaks, what I’m learning, and how small businesses can actually use AI without getting buried in hype.