Over the past 2 weeks I built an AI voice agent for a cosmetic clinic that operates across three locations in Amsterdam. The clinic often receives calls while treatments are in progress, which makes it difficult to answer consistently without interrupting patient care. So I built a voice agent that can: • check real appointment availability • create booking requests inside ClinicMinds CRM • verify bookings via SMS • route calls to reception when available • create callback requests when needed After two successful pilot builds, this project will become my first paying client deployment. The clinic is currently expanding and expects call volume to grow to around 20 calls per day in the coming months, so the goal is to have the system ready before that growth happens. The owner is very open to implementing new technology, which made this a great real-world use case to build and test with! Instead of building a demo, I focused on creating something that could actually be deployed in a real clinic environment. This is the walkthrough I submitted for the hackathon 👇 https://www.loom.com/share/122ea2b7649c4557affc4aad1a30c2bc Curious to see what everyone else built. Happy to answer a few questions if anyone is interested in how this system works!

No Mac Mini or VPS needed Just your local machine This is SpaceClaw Your personal AI assistant Similar to OpenClaw, but not a for No $100 to $200 setup cost Just $10 for an OpenAI API key Or use Gemini's free key Your API keys and personal credentials? Stored in a safe and separate environment Only you can see it What SpaceClaw can do: ➤Use different LLM models. ➤Remember past conversations with a 3-tier memory system. ➤Perform semantic search. ➤Take voice input and respond in voice. ➤Create, edit, delete, rename files. ➤Execute commands. ➤Perform web search. ➤Add or update its own skills. ➤Run with an MCP server. ➤Maintain a soul.md file that defines its personality and improve it over time. Grab it from here: https://www.notion.so/Space-Claw-315ccc22b8d380faa103d655dc9b7163?source=copy_link

Client problem: Manually sending the same onboarding email every time he adds a new web design client to Notion. His request: "Send an email when I categorize one element in Notion." My solution: Notion → Filters → Gmail (n8n workflow) Result: 15 minutes saved per client onboarding. Zero manual work. Another process automated. Here's the shift: Most people build complex workflows that look cool but solve nothing. That's why clients don't come back. I became an AI consultant who automates real business problems, one process at a time. Simple. Valuable. That's why clients keep coming back. Your turn: What's one repetitive task you want automated? Drop it below 👇 Got something? DM me. Let's automate it. #Automation #AIConsulting #n8n #BusinessAutomation #NoCode #WorkflowAutomation #NotionIntegration #ProductivityHacks

Most people use AI the same way: open a chat, type a prompt, hope for the best. Sometimes it works. Often it doesn't, because the AI has no idea how your business operates, what tools you use, or what good output looks like for your team. Agent skills change that. They're a simple, open format for giving AI agents the specific knowledge they need to do a job well. And they're about to reshape how marketing teams, designers, and agencies work with AI. https://dev.to/224industries/how-agent-skills-can-make-ai-actually-useful-for-your-business-249e

I recently built an automated outbound system that handles: - lead scraping - email cleaning + verification - short, personalized AI emails - pushing them straight into outreach campaigns Once set up, it runs daily with almost zero manual work. Sharing this because a lot of founders and freelancers here are still doing outreach the hard way. Happy to break down the workflow or answer questions if anyone’s curious.