Your whole team, AI-ready.
In one day.
A hands-on workshop where your team designs, builds, and deploys AI agents together. We bring the architecture. You bring the context. You leave with something live.
Not another AI 101.
This is for teams that already use AI and want to go further - building real AI-native capability, and leaving with something live. Not slides about what AI could do one day.
Non-engineering teams
Sales, Marketing, Ops, Support, People & Culture, Finance - teams that move the business but don't write code. This day was built for them.
Past the basics
Your team already uses Claude or ChatGPT. You want reliable agents, repeatable workflows, and real automation - not another prompting 101.
Ready to build something real
Not a slide deck. Not a framework. A working prototype, a prioritised roadmap, and a team that knows how to build more without needing Hourglass in the room.
Three phases. One day.
Every workshop follows this arc. The activities inside each phase are tailored to your team’s workflows, tools, and where you are on the AI maturity curve.
Break your role into agent-shaped work
Teams identify every repeatable task that could become an agent. We use the IPO framework to map inputs, outputs, frequency, and AI potential - from assist all the way to orchestrate.
Write instructions that make AI reliable
Move from prompting to proper agent briefing. Every group turns a real workflow into a production-grade brief: purpose, context, inputs, steps, output format, and guardrails.
From prompt to repeatable routine
Turn one-off AI work into automatic workflows. Time-based and event-based agents, human-in-the-loop design, smarter triggers. Each group leaves with a working operating model.
What you leave with.
Four concrete outputs from the day. Not ideas - artefacts you take home and act on.
Agent opportunity board
Every function maps its work into repeatable tasks, ranked by value, frequency, risk, and automation potential.
A well-briefed agent per team
Not just a prompt - a proper brief with purpose, context, inputs, steps, guardrails, output format, review loop, trigger, owner, and success criteria.
A live prototype
Each team builds or simulates a working agent using the tools they already have. Real, not theoretical - run before you leave.
A prioritised roadmap
A backlog of the next 5–10 agents to build, with dependencies, handoffs, and a clear path to a company running on connected AI workflows.
How a workshop day runs.
This is a structure that has worked well for teams we’ve delivered to. We adapt timing, sessions, and focus areas based on your team’s goals, existing AI maturity, and the use cases you most want to solve.
Opening
- 🧭Context - what AI-ready means for non-engineering teams
- 📊Framework - the AI maturity ladder: from chat to agent teams
- 🪞Reflection - where you are now, and where this day gets you
Demo block
- ⚡Live demos - advanced AI workflows running in production at Hourglass and in client companies
- 🤖See it working - our Slack agent, skills system, and business brain - live
- 🔥The goal - create genuine excitement for what's possible beyond basic prompting
MAP session
- 🗺️Identify - break every role into repeatable tasks that could become an agent
- 📈Rank - by AI potential: assist, delegate, automate, or orchestrate
- ✅Output - a comprehensive opportunity list for each function
BRIEF session
- ✍️Write - a strong agent brief for one real workflow per team
- 🛡️Cover - purpose, context, inputs, guardrails, quality bar, and edge cases
- 🛠️Set up - get set up with the tools you'll use to build
RUN session
- ⏰Triggers - time-based, event-based, and human-in-the-loop design patterns
- 💡Efficiency - smarter triggers to avoid wasting credits watching for nothing
- 🎯Tools - choosing the right tool for the right workflow
Build sprint
- 🏗️Build - teams race to build a real working agent using their actual tools
- 🤝Support - Hourglass facilitates, you bring the context
Showcase
- 🎤Present - agent name, problem solved, demo, trigger, and next steps
- ✨Inspire - see what's possible from a single building session
Next steps
- 🔗Connect - how agents hand work to each other across teams
- 🏢Architecture - company-wide AI architecture and shared context
- 🗺️Roadmap - prioritised agent roadmap, built together, taken home
Where this day gets you.
Most teams arrive at level 1 or 2 - using AI like a better search engine. A well-run workshop moves the whole team to levels 3–5 in a single day.
Claude Chat
Where most teams startYou ask, it answers. The output disappears when the tab closes.
Claude Cowork
Where most teams startYou share context before asking. Better outputs, still fully manual.
Skills
Workshop takes you hereSaved slash commands that run whole workflows consistently.
Shared Skills
Workshop takes you hereEveryone on the team runs the same skills. One quality bar, across the company.
Agents
Workshop takes you hereTriggers fire without you. Connected to your tools and data. No copy-pasting.
Agent Teams
The destinationAgents hand work to each other across functions without a human managing the handoff.
We show up with more than a deck.
Every workshop is facilitated by a Hourglass AI specialist. We demonstrate our own tools live - the same agents and workflows we run our business on - so your team sees what’s possible before they build.
Live demos from production
Our Slack agent, skills distribution system, and business brain - not slides about what AI could do.
Prepared for your use cases
We ask for your team's actual workflows and goals before we arrive. The day is built around those, not generic examples.
Hands-on during the build sprint
We work alongside each team during the build - not just lecture from the front.
Slack support after the day
Your team keeps building. We stay available to unblock them.
Anna Sekulic
AI Product Specialist, Hourglass AI
Anna has been building AI workflows since she was 16. She brings three years of hands-on experience designing agent systems for real businesses - the kind that run overnight, handle edge cases, and don’t need babysitting. She flies to you, runs the whole day in-person, and stays for questions at the end.
“I want every team to leave feeling like they built something real - not watched someone else build it for them.”
Anna Sekulic
Common questions.
30 minutes. Zero pressure. Total clarity.
Tell us about your business. If we’re a fit, your AI Audit can start the same week.