AI for Doers.
Understand the system. Direct it precisely. Put it on your real work.
- Duration 4 hours
- Audience ICs, knowledge workers, operations teams
- Format Live remote co-creation
- Prerequisite None
When to take this module.
- Your team is using ChatGPT in the shadows and you need a policy that fits how they work.
- Pilots keep launching and nothing sticks.
- Leadership wants a measurable productivity number and you need a path to evidence.
- Foundational literacy is missing and your team needs to practice on real tasks.
What the session covers.
01 Understanding the system
What an LLM actually is, what multimodality changes, and the Goldilocks rule for matching a model to a task.
02 Capability mapping and model fit
A taxonomy of what AI does well and where it breaks, mapped to your role. Heuristics for picking the right model by effort and capability tier.
03 Failure modes and responsible use
Five failure modes, three hallucination-catch tactics, and Green-Yellow-Red data classification.
04 Prompting on real work
A reusable Role-Task-Context-Constraints structure and an iteration loop. Applied across three use cases each participant nominates at intake.
What you walk away with.
- 01 A working mental model of how an LLM behaves and where it fails.
- 02 Three mode-tagged use cases drawn from your own job.
- 03 A reusable prompt structure that survives the next model release.
- 04 Working knowledge of responsible AI use at the individual-contributor layer.
Built for procurement, legal, and IT review.
Every participant leaves with documented EU AI Act Article 4 literacy training, attendance records, and an OpenBadge 3.0 credential your L&D can attest against.
LiftWork does not certify compliance on your behalf. We supply structured evidence to support your legal and L&D attestation.
Questions teams ask first.
01 Do participants need any AI background?
No. The module assumes plain-language users, not developers.
02 Do participants need to bring their own tasks?
Yes. The session runs on three real tasks each participant nominates at intake.
03 What if our company uses a sandbox account instead of personal logins?
That works. Sessions adapt to whatever AI surfaces are already approved on your stack.
Ready when you are.
A 30-minute call. We listen for the trigger pain, sketch the scope, and tell you whether the right shape is this module, another one, or neither.