Skip to content

otmanm/ai-systems-thinking-workshop

Repository files navigation

AI Systems Thinking Workshop

A practical 2-hour workshop that teaches non-technical professionals how to think and work with AI like a systems engineer.

No coding required. No prompt hacks. Just structured thinking, quality control, and reliable results.


Who This Is For

This workshop is designed for:

  • L&D managers building internal AI enablement programs
  • Product, operations, and marketing leaders using AI in daily workflows
  • Consultants and managers who need reliable AI outputs, not demos

Delivered to 700+ professionals across EMEA (Startups & Fortune 500) with 73% sustained adoption rates.


What Participants Learn in 2 Hours

By the end of this session, participants will:

✅ Frame AI tasks using context, constraints, and intent - not "better prompts"
✅ Diagnose why AI outputs fail, drift, or produce generic results
✅ Apply 4-Q thinking to validate AI recommendations
✅ Use simple quality control checklists before shipping AI work
✅ Decide when AI should assist and when humans must intervene


Workshop Structure

Time Module Activity Deliverable
0:00-0:15 Why AI outputs fail Group diagnosis of generic AI output Mental model reset
0:15-0:40 Context vs. prompting Theory + live demo Context brief template
0:40-1:15 Hands-on practice 3 exercises with real scenarios Before/after examples
1:15-1:45 Quality control Output review + 4-Q validation QC checklist
1:45-2:00 Transfer to work Action planning Personal use case

Delivery formats:

  • Live in-person (recommended)
  • Live remote (Zoom/Teams)
  • Internal enablement session

View detailed 2-hour agenda →


What's Included

📚 Workshop Materials

🧠 Theory Modules

  • Context vs. prompt - Why longer prompts often fail
  • Human bottleneck analysis - Control design for AI systems
  • AI readiness audit framework - 14-day pilot methodology
  • Readiness Gate for non-technical operators - The 21-section field guide that explains why AI plans look impressive and then fail in real execution. Covers the planning fallacy, the cooking & iceberg model, plan A/B/C/D without wasting resources, evidence labels, examples of what goes wrong, and an embedded workshop facilitation script. Written for interns, operators, project managers, founders, and assistants who need to use AI for serious planning without getting fooled by confident nonsense.

✏️ Hands-On Exercises

  • Exercise 1: Context framing - Turn vague requests into structured briefs
  • More exercises coming: Before/after improvement, Quality control

📋 Templates & Checklists

🔍 Before/After Examples

🛠️ Prompt Libraries


Quick Start

For L&D Teams

  1. Review 2-hour agenda
  2. Test Exercise 1 with your team
  3. Print context brief template as handouts
  4. Run pilot session with 8-12 participants

For Individual Learners

  1. Read Context vs. prompt
  2. Complete Exercise 1
  3. Apply context brief template to your next AI task
  4. Use QC checklist before shipping outputs

For Hiring Managers

Evaluating AI training capability? Check:

For Operators & Interns

You've been asked to use AI for a real plan, deliverable, or decision and you want to do it right:

  1. Read the Readiness Gate end to end (45 minutes — it's long because the failure modes are many)
  2. Apply the One-Page Reminder before any AI-generated plan, roadmap, build, or campaign
  3. Use the Minimum Viable Preparedness Formula (section 18) to decide whether a plan is even allowed yet
  4. When you do generate a plan, run it through the Prompt QC Checklist before shipping

Companion Repositories

This workshop is the non-technical foundation. Two sibling repos go deeper:

  • ai-evaluation-workshop — 20-module mastery curriculum on measuring, testing, and shipping production AI systems (Ragas, DeepEval, LangSmith, FutureAGI, Promptfoo, CI/CD quality gates, red-teaming).
  • voice-agent-deployment-kit — 39-module deployment kit for production AI voice agents and omnichannel messaging with SLAs, GDPR/AI Act compliance, and vertical playbooks for EU markets.

Together, the three form a complete ladder: think → measure → deploy.


Real Results

Fortune 500 Tech - Multilingual Sales Enablement

  • 400+ sales reps trained across 18 EMEA markets (EN/FR/ES/AR)
  • 73% sustained adoption rate (industry baseline: 45%)
  • Scaled from 4 to 18 languages in 12 months
  • Renewed annually due to measurable ROI

SaaS Startup - Operations Workflow Optimization

  • 25-person ops team, 14-day pilot
  • 30% reduction in approval latency
  • Quality maintained (rework rate stable)
  • Framework embedded in 3 departments

Making This Workshop Better

Current development priorities:

  • Executive briefing version (60 minutes)
  • Customer support workflow examples
  • Healthcare context examples
  • Multilingual exercise variants (FR/ES/AR)
  • L&D evaluation rubric with pre/post assessment

See all open projects →

Contributions welcome - open an issue with your adaptation needs or submit a PR.


Author

Otman Mechbal | AI Educator & Automation Strategy Advisor | Teaching teams to avoid AI slop & over-automation through smart methods | 700+ pros trained | Startups to Fortune 500


License

MIT License - Use commercially, adapt for your organization, share with attribution.

If this workshop prevents even one failed AI pilot, it's worth it.

About

A practical workshop that teaches non-technical professionals how to think and work with AI like a systems engineer.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages