In four live weeks, you build and ship real AI agents that take real work off your plate, run reliably, and are ready to use with your team. No coding background required.
“I built a Chief of Staff agent, an automated newsletter, and an eBay agent…”
“It changed EVERYTHING!”
“You build, you break things, you ship.”
You run a team or a function, and everyone looks to you for the agent answer. You've read all of it. But you've never shipped one yourself, and leading a shift you've never made is wearing thin.
You're a consultant, PM, marketer, or ops lead. You live in these tools, but it stops at chat. You want agents that take the repetitive work off your plate, and the proof to put them in front of your team.
You're a VP, director, or founder in 50–60-hour weeks. You've been "AI-curious" for a year and still mostly use it to rewrite emails. A fixed schedule, a hard deadline, and a room of peers are how it finally happens.
Workflows that run with you in control, products you can show, and a cohort of builders who keep each other sharp. Agents execute. You retain control.
An agent that scans your inbox, calendar, and the web overnight, then hands you a personal briefing every morning. It runs on your schedule; you decide what it acts on.
Tools, memory, and agents wired into one system your whole workday runs through, not a drawer of disconnected chatbots.
Your own product, live at a public URL by Demo Day: shareable, usable, and ready for your team to use.
A coordinated team of agents with defined roles, hand-offs, and clear review points that keeps work moving while you stay accountable.
A private Agentic AI Cohort community of operators and builders: feedback and accountability that outlast the four weeks.
100+ pre-built agent packages, plus the prompts, workflows, and templates you build along the way - yours to keep long after the four weeks.
8 live build sessions + 3 tech clinics. Two live build sessions a week — Tuesdays & Thursdays, 16:00 UTC, 60 min, all recorded — plus a Friday tech clinic in weeks 1–3, so nothing blocks your build. Runs Sep 1–24, 2026.
The mental model first: how these agents actually think, and the dials you control - context, memory, tools, skills. Then your technical setup (tools, accounts, community) and your first agent: an AI chief of staff that works for you every morning.
Your full setup: tools, accounts, community access. Nothing blocking you from building.
Andreas walks you through his actual morning, run by AI agents. Then you build your own: a chief-of-staff agent that pulls from your accounts, scans the web, and delivers your personalized daily briefing. Automatically.
Troubleshooting only, no teaching. Get unblocked with Andreas' colleagues, so everyone stays operational.
From one agent to a system you can rely on: context, memory, tools, skills, multi-agent workflows - and the layer everyone skips: evals, guardrails, cost.
A hands-on terminal deep dive and the basics of agentic engineering, so the technical layer stops being a black box.
CLAUDE.md, memory strategy, custom skills - repeatable playbooks it runs on command - and your agent's "constitution". Align it with your goals and context so it works the way you do.
Troubleshooting only, no teaching. Get unblocked with Andreas' colleagues, so everyone stays operational.
A special guest takes you inside an app with paying customers, then you learn the full stack to take your own build live and charge for it.
A senior Big Tech engineer walks through an app he shipped that makes money every month: how it came together, what broke, and how to build more complex workflows yourself. Guest announced to enrolled members.
The full deployment stack (Supabase, Vercel, Claude Code, Lovable) and how they fit together. Take what you've built, put it live on the internet, share it with friends or teammates, and set up a payment page so you can charge for it.
Troubleshooting only, no teaching. Get unblocked with Andreas' colleagues, so everyone stays operational.
Scale from one agent to a team of them working for you, and ship your capstone. Demo Day caps the cohort. You leave with a coordinated agent team that keeps work moving while you remain accountable.
Go from a single agent to a coordinated team: a researcher, a builder, a reviewer that hand off to each other while you direct. Then get your capstone demo-ready.
Show your best build to the whole cohort: live demos, direct feedback from Andreas, and a proper send-off. You leave with something that runs.
Andreas Horn is VP AI & Growth at BLP Digital and the author of Human in the Loop. Former IBM and Microsoft, he has spent 12+ years building and scaling AI systems and has delivered dozens of AI projects for enterprises around the world, including Fortune 500 companies.
Today he builds AI-native operations at BLP Digital and advises leadership teams on moving AI from pilots into production. Before that he led the AIOps service portfolio across EMEA at IBM, contributed to its global AI Agents offering, and worked on enterprise AI at Microsoft. He builds with these tools every day, which is why the Cohort teaches what works, not what is announced.
Today, Andreas is one of the most-followed voices in enterprise AI: 300,000+ followers across platforms and 38,000+ newsletter subscribers. He is the author of The Generative AI Career Masterplan and The Definitive Guide to Claude Code (both Packt), a featured AI instructor on platforms such as LinkedIn Learning, and a regular lecturer in master's programs focused on artificial intelligence.
He is also a sought-after guest lecturer at universities around the world, including UC Berkeley, Hult International Business School, Shanghai Jiao Tong University, and United Arab Emirates University.
assets/andreas-stage.jpg
assets/andreas-keynote.jpg
assets/andreas-4.jpgBy the end you've shipped working agents and demoed them to the room, while most people in your company are still in the pilot-deck phase.
“The people who build and manage AI agents are becoming the most valuable people in every organization. Almost nobody in your organization is one of them yet.”
Andreas HornFounder, Human in the Loop
Every seat includes partner credits for the tools you build on. Team of 3+? Email for team seats. Whole department? That is HITL Advisory.
That's exactly who this is built for. Consultants, VPs, executives, ops leads: people who make decisions about AI and need to understand it from the inside, not just talk about it. You'll build the agents yourself. And once you've shipped one, you're not selling your team a strategy you've never run.
Yes. The cohort is designed as practical professional development for leaders, operators, consultants, and teams building with AI agents.
We provide a company-ready invoice and an employer approval letter to help you submit the course through your learning & development, training, or professional development budget.
For self-employed professionals and businesses, the course may also be relevant as professional education or training expense. Please check the individual tax treatment with your accountant or finance team.
Bringing 3+ people? Email us.
No coding and no AI background needed. What you do need: genuine curiosity about how this actually works, not just what it can do. A real reason to learn: something you want to build, or a problem you're tired of working around. And the willingness to sit with confusion for a bit and push through it. That's it.
The program is 8 live build sessions + 3 tech clinics across four weeks. About 5–8 hours a week: two live build sessions, a weekly project, and some async reading. Structured, but not overwhelming if you're organized. And if a session doesn't fit your calendar, every one is recorded. Recordings and resources stay available to you. The course is built for working professionals. That's the whole design.
Simple: try the first week, risk-free. If it's not for you, email us any time in the first week and we'll refund you in full. No questions, no forms.
Cohort 2 runs Sep 1–24, 2026. 8 live build sessions + 3 tech clinics, a room of peers — and a week-one full refund if it's not for you.
Enroll nowHuman in the Loop helps business leaders and executives move their organizations from AI experimentation to AI operating capability.