Claude AI agent workshops & training for mid-market teams
In half a day, your team goes from curious about AI to shipping agents that do real work, built on the tools and workflows you already use. No slideware, no homework, no generic demos, just a Claude AI agent workshop that ends with agents in production.
- ½
day to working agents
- 6
production-ready agent
- 10
Claude surfaces
What does your team walk away with?
A Claude AI agent workshop should send your team back to work with things they keep, not notes. Here is what each person and team takes away.
Working agents, live before you leave
Your team builds and tests real agents during the session, working in three pods of two, for six production-ready agents scoped to your actual week: meeting prep, request triage, candidate comparison, or whatever fits. Each one is running on your real tools before the session ends.
A reusable skills library
Custom SKILL.md files that teach Claude your processes once and apply them to every future task. Contract review, report formatting, a house style for briefs: written as Skills your team owns and keeps extending long after the workshop. This is the asset that compounds.
Role-specific playbooks
Each attendee leaves with a personal playbook tailored to their function, not a generic handout. It names the agents worth building next for their role, the connectors those agents need, and the prompts that already worked.
Shared Claude fluency
Your whole team leaves able to tell a Chat prompt from a Project, a Skill from a Managed Agent, and to know which surface a task calls for. No one has to be talked through the basics again, and no prior AI experience is assumed going in.
A map of your automatable work
The discovery sprint leaves you with a ranked list of the tasks worth automating first, clustered into agent opportunity zones. It is a backlog your team can keep working on long after the workshop.
A prioritized 90-day roadmap
A clear plan for the next quarter, plus a graduation path. When an agent outgrows the Claude platform and needs enterprise reliability or governance, it moves to our AI agent development services. When the backlog is bigger than one team, it feeds an AI consulting engagement.
How does a Claude AI agent workshop work?
Working agents, a reusable skills library, role playbooks and a 90-day roadmap: that is what your team walks out with. Here is how the half-day itself runs, segment by segment.
Frame & foundations (45 minutes)
We orient the team on all ten Claude surfaces, so everyone shares a mental model before building. No one writes a prompt until the whole room can name the tool for the job.
Brainstorm & cluster (30 minutes)
A guided discovery sprint where your team surfaces the tasks it repeats most, the handoffs that break, and the context that gets lost. We cluster those into agent opportunity zones and pick the ones worth building today.
Build hour (2 hours)
Your team splits into three pods and builds two agents each, six in all, on its own workflows, with a facilitator at every table. Not a demo, not a sandbox: agents connected to your actual tools, running before the hour is out.
Share-back & next steps (30 minutes)
Each pod demos what it built, we pressure-test it together, and you leave with a ranked backlog and a 90-day roadmap for what to build next.
What's covered in Claude training?
Your team walks out fluent in all ten Claude surfaces, knowing when to reach for each one. That is the Claude AI training you cannot get from the product docs.
- Chat: The starting point for any task.
- Connectors: Gmail, Drive, Slack, Notion and 50+ more, via Claude AI integration.
- Projects: Context that travels with the work.
- Artifacts: Docs, slides, apps and dashboards.
- Skills: Teach once, apply forever.
- Cowork: Claude works on your files directly.
- Claude Code: Terminal-native for engineering.
- Managed Agents: Hosted, long-running workflows.
- Claude Design: Visual work through conversation.
- Claude in Chrome: Browser-native agent automation.
Who are Claude AI agent workshops for?
This is corporate AI training scoped to one team's actual week, not an all-hands literacy session. It works best for a focused group of 6 to 15 people who share workflows. Three functions get the most out of it.
Operations and finance teams
Replace repetitive coordination, reporting, and triage with agents that know your workflows and connect to your stack, the same pattern behind our AI automation services and the scheduling portal we built for Apex Leaders. Inbox and request triage. Weekly briefings and status summaries. Contract review and compliance flagging. Meeting prep pulled from calendar and email.
HR and recruiting teams
Hand the administrative load to agents, so your team can focus on the areas that need a person’s touch. In the session an HR team will decide on what to build. For example, they can build agents that compare candidates against a rubric, run onboarding workflows and checklists, answer policy questions from your handbook, and schedule interviews. The hiring judgment stays with your recruiters; the admin runs on agents.
Engineering and product teams
Put Claude to work inside the build, not just around it. Claude Code for real changes. Managed Agents for long-running jobs. MCP connector development for the tools you use. Code review triage.
Frequently Asked Questions
What is an AI agent workshop?
An AI agent workshop is a structured, hands-on session where a business team learns to build and deploy AI-powered agents on their own workflows. LaunchPad Lab’s workshops are half-day sessions using Claude, covering the full Claude toolbox across all ten surfaces, a discovery sprint to identify the highest-value use cases, and a build hour where real agents are created and tested before the session ends. Unlike general AI training, teams leave with working output, not slide decks.
What AI agents does a team typically build in the workshop?
Teams build six production-ready agents in a single session, working in three pods of two. Common examples include a Meeting Prep Agent (pulls context from Gmail and Calendar), a Request Triage Agent (routes incoming asks to the right queue), and a Candidate Comparison Agent (stacks CVs against a rubric). The specific agents are always determined during the discovery sprint, scoped to that team’s actual week, not generic demos.
Who are AI agent workshops designed for?
LaunchPad Lab’s workshops are designed for specific business functions, not general audiences. The most common teams are operations and finance (triage, reporting, meeting prep), HR and recruiting (candidate comparison, onboarding, policy Q&A), and engineering and product (Claude Code, Managed Agents, MCP connector development). Workshops are not appropriate as all-hands AI literacy sessions; they work best with a focused team of 6 to 15 people who share workflows.
Is there a path to custom AI products after the workshop?
Yes, every workshop ends with a prioritized agent backlog and a clear graduation path. Agents that outgrow the Claude platform, because they need enterprise-grade reliability, multi-agent coordination, or IT governance, can graduate to LaunchPad Lab’s AI agent development services, where we build the custom product. The workshop is explicitly designed as the first step in that continuum, not a standalone event.
What does a Claude training course cover, and how is a workshop different?
A Claude training course usually walks an individual through the features of Claude on a set curriculum, and you leave knowing what the tool can do. A LaunchPad Lab workshop is the opposite trade: a team, its own workflows, and working agents built and deployed before the session ends. A course builds fluency, a workshop builds output. The workshop assumes no prior course.
How long does an AI enablement workshop take?
The core AI Enablement Sprint takes half a day, approximately 3.5 hours. The workshop moves through four structured segments: Frame & foundations, Brainstorm & cluster, Build hour, and Share-back & next steps. Full-day cross-functional workshops covering two teams are also available, as are multi-week enterprise programs for org-wide rollout.
What tools and integrations are covered?
The workshop covers the full Claude toolbox: Chat, Connectors (Gmail, Google Calendar, Google Drive, Slack, Notion, and 50+ more via MCP), Projects, Artifacts, Skills, Cowork, Claude Code, Managed Agents, Claude Design, and Claude in Chrome. The session also covers the Model Context Protocol (MCP) open standard, which is what allows Claude to talk to your existing stack, and how engineers can build custom connectors when an existing one does not exist.
What's the difference between this and standard AI training?
Standard AI training is informational; teams leave knowing what AI can do. LaunchPad Lab’s workshops are output-oriented; teams leave with working agents deployed on their actual workflows. The key differences are the structured discovery sprint (which maps specific high-friction tasks to agent opportunities), the hands-on build hour with real facilitation, and the custom Skills library and role-specific playbooks each attendee receives to keep building after the session.
Where is LaunchPad Lab based, and do they run remote workshops?
LaunchPad Lab is headquartered in Chicago, IL, and runs AI agent workshops for teams across the United States. Workshops can be facilitated on-site at your location or delivered remotely via video conference; the format depends on team size, preference, and scope.
How much does an AI agent workshop cost?
Cost depends on scope: how many teams, the format (a half-day single-team sprint, a full-day cross-functional session, or a multi-week enterprise rollout), and whether you want post-workshop support. We align on the right format and gives you a clear number before you commit to anything.
Ready to build with your team?
Tell us which team, what they do, and when you want to run it. We'll scope the right format and get it on the calendar.