The new Agent Operators episode with Karsten Stinis is live. Karsten installs managed AI agents for businesses in Belgium. His offer is €2,750 for the setup, then €250 a month for support. The part I'd pay attention to is what happens during the installation: he watches the client work, finds a recurring task, and aims to have something useful waiting for them the next morning. Before you install the agent, decide what it will finish. A morning briefing is a good example: meetings, open tasks, and unanswered emails worth following up on. The person can open it, check the sources, and decide whether it helps. Whether you're building for yourself or a client, try scoping one overnight job: 1. The job: choose one recurring task. 2. The inputs: name the inbox, calendar, or task list it needs. Connect the accounts and check access. 3. The result: decide where the output should appear and when it should be ready. Make it easy to trace back to its sources. 4. The review: choose who checks it. Keep outgoing messages as drafts for the first run, and have a person approve them. Run it once. Read the output with the person who will use it. Use their corrections to improve the next run before adding another job. In the episode, Karsten walks through Hermes on Orgo, Composio for connecting business tools, Honcho for memory, and how he monitors usage and incidents. Keeping the workflow working is part of the service. He also reports one client's agent setup being built in 22 minutes. Humans still handle the tool and account connections. Watch the full conversation: How He Sells AI Agents for €2,750 + Monthly Support https://www.youtube.com/watch?v=Y8MJG3iDtSo What would you want your agent to have ready tomorrow morning? Share the job below.

I just built a full AI employee from scratch in this video. Everyone’s hyped on agents but barely anyone’s actually showing how to build ones that can really work inside a business. Here’s the entire stack I spun up live (no holding anything back): • Cloud computer running Hermes on Orgo • Its own email (AgentMail) • Its own phone number (AgentPhone + iMessage) • Telegram as the main chat interface • Every tool & connector hooked up through Composio • Credit card so it can actually spend money • Obsidian vault as its knowledge base / second brain • Latitude for observability so I know when shit breaks I made one main orchestrator agent (named it Hubert) that stays in charge 99% of the time and just hands tasks off to specialized sub-agents underneath it. One giant bloated agent is a nightmare to debug. This way everything stays clean and purpose-built. Full live build, every prompt I pasted, every terminal I opened… it’s all in there. And I’m giving away the complete templates for the orchestrator + every sub-agent I build going forward. I'm giving you actual employees that can do real work. Watch the whole thing here: PS - the template is free here on Github: https://github.com/nickvasilescu/nicks-stack

I'm building a bunch of one-click templates with everything pre-installed and initialized for your agent of choice relevant skills, relevant connectors, plugins, etc. what are some template one-click deployment agents that you would like to have? examples might include a Hermes agent spun up with Composio, agent mail, agent card, etc., pre-configured and ready to go any ideas?

The model: I sell managed AI agents to businesses for $5K/month each. I handle the infrastructure, they get an employee that never sleeps. 10 clients puts you at $50K MRR with 85%+ margins, run entirely by you and a fleet of agents. Here's the playbook: 1. The arbitrage is that nobody knows this is possible. 99% of business owners are still asking ChatGPT what the weather is. One working agent hooks them on the spot. 2. Sell abundance. Unlimited agents, unlimited infrastructure. They don't care what an MCP is, they care that their problem is gone. 3. Don't niche too early. Say yes to everyone and let the market pull you. You find the niche by doing reps, not guessing. 4. Paid audit into managed service. Charge $1K to map every automation opportunity, then credit it toward month one. It qualifies the lead and makes the upsell a no-brainer. 5. First call, don't sell. Record it, map the workflow tip to tail, find the automation with the most value and least effort. Start there. 6. My stack is Hermes + Composio + Orgo. Composio connects all their apps in one click. Orgo spins up a working Hermes agent in 26 seconds. 7. Productize with a golden snapshot. Build one perfect agent, clone it, and every copy comes over one-for-one with auth intact. 8. Turn client call transcripts into skills in 10 minutes. Feed the recording to Claude Code, write the skill, port it to the client's agent via Orgo MCP. 9. Watchdogs make you look elite. Get alerted before the client notices anything broke. "Already fixed it" is why they keep paying you. 10. You become their guy. You drive more outcomes than their own employees. They credit every win to you, and churn drops to almost nothing. My 2 biggest takeaways: 1. Bet on cost going to zero. We launched unlimited tokens when it was barely profitable because we knew we'd capture the spread. Build for where the puck is going. 2. One client every six weeks gets you to $600K a year.

November 2025, I had one month of runway left. I was living in a hacker house in Berkeley with 15 other people, working out of an electrical room under the stairs, trying to build my AI agent business at the time called ComputerUse.Agency. Before that, I had been running an AI edtech platform that was doing a few grand a month at one point, but by then it was trending to zero. So I had a choice. Keep trying to save the thing that was slowly dying, or go all-in on something bigger. My now co-founder, Spencer, was building the early days of Orgo, and I thought to myself: “What the hell? Let me try selling computer use agents to businesses as AI employees.” The next morning, I booked seven calls. By the end of the week, I had my first paid customer: a $3,500 implementation for an AI agent automating distribution workflows. And the whole thing was built on top of Orgo. Orgo was originally meant to be more of an AI researcher tool for computer use agents, but I was using it differently. I was running a business on top of it. At some point, Spencer and I looked at each other and realized I had more revenue than he did. We were like, what are we doing? We should just work together. Fast forward to January 24th. Spencer was landing in San Francisco, and I was supposed to pick him up from the airport. I was already running late, but there was this new thing called Clawdbot. Out of pure excitement, I pulled out my camera and made a quick video explaining what Clawdbot was and how to set it up on Orgo. The first 10 minutes, it was crickets. At the time, I had no following. I would get one or two likes on most tweets. Truly, nobody was watching. But then the video started moving. It went on to do 1.5 million impressions on Twitter, then another million on Instagram, then another million on TikTok, YouTube Shorts, and everywhere else. That post was the first video to ever go viral for Clawdbot, or OpenClaw, as it’s now known. To this day, I’m still on the front page of OpenClaw’s website, slight flex, I know.