LLaManchaAI Enablement
Checking access…
← All modules
Module 0Audience: all

Orientation and kickoff

Sets expectations for the company rollout, explains how completion works, and frames AI as a practical workflow capability.

Outcomes

What you will be able to do

  • Understand why the company is investing in AI enablement.
  • Know what completion requires and what data not to submit.
  • Pick one personal workflow goal for the program.
  • Set a baseline you will re-rate at the end of the program.
Completion check

How this module is approved

State your role, one workflow you want to improve, one boundary, and your five baseline self-ratings.

Pass criteria

  • • Role is named
  • • Workflow goal is specific
  • • Boundary references data, quality, or review
  • • Baseline self-ratings present
Loading personalization context…
Lesson20–30 minutes self-guided, or 30–45 minutes with a kickoff discussion

What you should take away

By the end of this orientation, participants should understand that this program is not a generic AI lecture. It is a practical enablement path for using AI safely on real work while keeping human accountability, company rules, and data boundaries intact.

Part 1

Why this program exists

AI tools are already showing up in everyday work: drafting, summarizing, coding, researching, documenting, planning, and troubleshooting. The risk is not just that employees use AI. The bigger risk is that everyone improvises alone, with different assumptions about what is safe, useful, allowed, and good enough.

  • Give employees a shared baseline for responsible AI-assisted work.
  • Move from random experimentation to repeatable workflows.
  • Help teams identify where AI can save time without weakening quality or trust.
  • Create reusable role, workflow, and policy artifacts the company can build on after training.
Activity · ~3 minclassify scenario

Before the program teaches you anything, capture a baseline. You will re-rate these same five dimensions at the end, so be honest now — this is for you, not a grade.

  • Scale: 1 = never done it · 3 = use it but no method · 5 = methodical and safe
  • Rule: Rate from memory in 60 seconds; do not overthink

Your task

Rate yourself 1–5 on each: (1) writing a prompt with enough context, (2) spotting a likely-wrong AI answer, (3) knowing what data must never go into a tool, (4) building a small artifact with AI, (5) deciding when to escalate to a human. Write the five numbers down and keep them.

Show a hint
If you are unsure between two numbers, pick the lower one — it makes the end-of-program comparison more honest.
Compare with a strong answer
A common honest baseline early in a program is something like 3 / 2 / 3 / 1 / 2 — comfortable using AI casually, weak at verification and artifact building. There is no 'right' answer; the value is the delta you will see later.

Why this matters: Retrieval and self-assessment at the start create the bookend that makes growth visible at the end. A program with no baseline can only ever feel useful.

Quick check · <30 sec

What is the bigger organizational risk this program addresses?

  • A. Employees using AI at all
  • B. Everyone improvising alone with different assumptions about what is safe and good enough
  • C. AI being too slow
  • D. Not enough AI tools
Show answer
The risk is not usage; it is uncoordinated usage. A shared baseline is the point of the program.

Part 2

What this program is — and is not

This is workplace enablement, not a formal certification program and not a promise that AI can replace judgment. The goal is to help people use AI as an assistant: a tool for acceleration, drafting, comparison, explanation, and review support. The employee and the company still own the decision, the quality bar, and the final output.

  • AI can assist with drafts, outlines, explanations, examples, refactors, and first-pass analysis.
  • AI should not be treated as an authority, source of truth, confidential data vault, or unsupervised decision-maker.
  • Completion means the participant submitted useful evidence of learning, not that LaMancha is licensing or certifying the person professionally.

Quick check · <30 sec

True or false: completing this program certifies you as a professional AI practitioner.

  • A. True
  • B. False
Show answer
False. Completion is evidence of learning on real workflows. The employee and company still own every decision and the quality bar.

Part 3

A 60-second framing you can reuse

If a teammate asks 'what is this AI thing, really?', this is the script. It is deliberately short and unhyped — the calm version is the accurate one. (You may reuse this as a kickoff talk track; production of any video is out of scope for the program itself.)

  • Script: 'AI is a fast, fluent assistant that is sometimes confidently wrong. It is great for first drafts, summaries, and comparisons on low-risk work. It is not a source of truth, a vault for sensitive data, or a decision-maker. You stay accountable: you give it sanitized context, you check what it gives back, and you keep a human at anything that affects someone else.'
  • Notice what the script does not say: nothing about transformation, revolution, or replacing people.
  • The honest framing is also the one that ages well.

Quick check · <30 sec

Which phrase does NOT belong in an accurate 60-second AI framing?

  • A. Sometimes confidently wrong
  • B. You stay accountable
  • C. It will transform everything
  • D. Great for low-risk first drafts
Show answer
'It will transform everything' is hype, not information. The calm, specific framing is the accurate one.

Part 4

How completion works

Each module ends with a practical completion check. Participants submit a short artifact: a reflection, classification, prompt/output example, role persona, workflow brief, or policy-readiness input. Submissions should be specific enough to show learning, but sanitized enough to protect the company.

  • Use realistic but sanitized examples.
  • Do not paste customer, patient, financial, credential, source-code, legal, HR, or proprietary data unless the company has explicitly approved that tool and that use case.
  • When in doubt, describe the scenario generically instead of pasting the underlying sensitive content.
  • The best submissions show judgment: what AI helped with, what the human checked, and where the boundary is.

Quick check · <30 sec

You want to show a real workflow but it involves customer data. What do you submit?

Show answer
Describe the scenario generically — name the data category, not the data. A sanitized description that shows your judgment scores higher than a real-but-leaky one, which is returned for revision.

Part 5

Your first AI workflow goal

Start small. A good first AI workflow is frequent, low-risk, reviewable, and annoying enough that saving time matters. The program works best when each participant picks one concrete workflow they can improve during the training path.

  • Good starter examples: summarize meeting notes, draft a status update, rewrite technical notes for a nontechnical audience, create a test checklist, compare options, or generate first-pass documentation.
  • Poor starter examples: make employment decisions, evaluate medical/legal/financial questions without expert review, process confidential client records in an unapproved tool, or ship AI-generated code without tests and review.
Activity · ~5 minclassify scenario

You will sort six candidate first-workflows into 'good starter' or 'not yet', then pick yours. The point is to choose deliberately, not by enthusiasm.

  • Test: Frequent · low-risk · reviewable · annoying enough to matter
  • Cards: (a) weekly status draft (b) screening job applicants (c) summarizing your own meeting notes (d) interpreting a customer's contract (e) drafting a checklist (f) deciding incident severity

Your task

Place each of the six on 'good starter' or 'not yet' and say why. Then name your own first workflow and confirm it passes all four parts of the test.

Show a hint
Anything that decides something about a person, or needs expert review to be safe, is 'not yet' for a first workflow.
Compare with a strong answer
Good starter: a (frequent, reviewable, low-risk), c (low-risk, you own the source), e (low-risk, reusable). Not yet: b (decides about a person), d (needs legal expertise), f (high-stakes operational judgment). My first workflow: weekly status draft from sanitized bullets — frequent, low-risk, reviewable, and currently tedious.

Why this matters: Choosing the first workflow against an explicit test is the difference between a win you can measure and a demo you forget.

Quick check · <30 sec

Which disqualifies a workflow as a good first pick?

  • A. It happens every week
  • B. You can easily tell if the output is wrong
  • C. It makes a decision about a person or needs expert review to be safe
  • D. It is mildly annoying
Show answer
Frequent, reviewable, mildly annoying are all good signs. Deciding about people or needing expert review pushes it to 'not yet'.

Part 6

After this module: continuous practice

Orientation is not the work; it is the on-ramp. The participants who get value treat the program as a loop: pick the workflow, apply the next module's idea to it, keep the persona and prompts you build, and re-rate your baseline at the end. Skill here decays without reps, exactly like any other.

  • Carry one workflow through the whole program rather than starting fresh each module.
  • Keep every artifact you build (persona card, prompt, workflow brief) — they compound.
  • Re-rate the five baseline dimensions after the final module and compare.
  • After the program: schedule a recurring 15-minute review to refresh prompts and personas as your role drifts.

Quick check · <30 sec

What is the single best way to get value from the rest of the program?

Show answer
Carry one concrete workflow through every module and keep the artifacts you build, instead of treating each module as a standalone lecture.

End-of-module quick check

Five short retrieval questions. Answer from memory first, then reveal each explanation.

  1. 1. This program is best described as…

    • A. A professional certification
    • B. Workplace enablement: using AI safely on real work with humans accountable
    • C. A tool license
    • D. A one-day workshop
    Show answer
    It is practical enablement on real workflows, not certification, and accountability stays with the employee and company.

  2. 2. A good first workflow is…

    • A. Rare and high-stakes
    • B. Frequent, low-risk, reviewable, and annoying enough to matter
    • C. Anything customer-facing
    • D. Whatever is most exciting
    Show answer
    The four-part test keeps the first win measurable and safe.

  3. 3. True or false: if a real example involves sensitive data, you should paste it so the submission is realistic.

    • A. True
    • B. False
    Show answer
    False. Describe it generically by data category. Realistic-but-leaky submissions are returned for revision.

  4. 4. Why capture a baseline self-rating now?

    • A. It is graded
    • B. So growth is visible when you re-rate at the end
    • C. To compare against coworkers
    • D. It is required by law
    Show answer
    The baseline is the bookend that makes end-of-program growth measurable instead of a feeling.

  5. 5. Name the single best habit for getting value from the rest of the program.

    Show answer
    Carry one concrete workflow through every module and keep the artifacts you build, rather than treating modules as standalone lectures.

Further reading

Worked examples by role

Front office coordinator

Orientation response: front office coordinator

I coordinate scheduling and intake. The workflow I want to improve: turning the day's sanitized voicemails into a prioritized callback list. Boundary: no personal or contact details go into any tool that is not the approved system, and the front-desk lead reviews the list before the team works it. Baseline: 3/2/3/1/2.

Sales coordinator

Orientation response: sales coordinator

I support the sales desk. Workflow to improve: first-draft follow-up emails from sanitized account context. Boundary: no customer identifiers, deal values, or competitor claims into an unapproved tool; the account owner sends after fact-checking. Baseline: 4/2/2/1/3.

Software developer

Orientation response: software developer

I am on a delivery team. Workflow to improve: turning sanitized incident notes into a clear follow-up action list. Boundary: no production secrets, customer data, credentials, or proprietary source into an AI tool; I review any summary before sharing. Baseline: 4/3/4/3/3.

Completion artifact

Submit your role, one specific workflow you want to improve during the program, one boundary you will follow, and your five baseline self-ratings.

ExercisePilot-ready artifact

Use this orientation exercise to anchor the rest of the program in one realistic, low-risk workflow. Keep the answer short, specific, and sanitized.

Participant template

  • Role or team:
  • Workflow I want to improve:
  • Why this workflow matters:
  • AI boundary I will follow:
  • Human review or approval gate:
  • Baseline self-ratings (5 numbers):

Example submission

Role or team: software delivery lead. Workflow: turn sanitized incident notes into a follow-up action list. Why it matters: it saves time after recurring support events. Boundary: no production secrets, customer identifiers, credentials, or proprietary source code go into an unapproved tool. Review gate: I review the summary before sharing it with the team. Baseline: 4/3/4/3/3.

Role-flavored variants

Same exercise, framed for different roles. Use the one closest to your work.

Sales coordinator

Pick a customer-facing workflow. The boundary must name customer identifiers and competitor/pricing claims; the review gate must keep the account owner on the send.

See a sample submission
Sales coordinator orientation: Role: sales coordinator. Workflow: first-draft follow-up emails from sanitized account context. Why: drafting is the slow part. Boundary: no customer identifiers, deal values, or competitor claims in an unapproved tool. Review gate: account owner fact-checks and sends. Baseline: 4/2/2/1/3.

Training manager

Pick a curriculum or evaluation workflow. The boundary must name learner records; the review gate must include a bias/accessibility check.

See a sample submission
Training manager orientation: Role: training manager. Workflow: first-draft comprehension questions from objectives. Why: drafting volume is the bottleneck. Boundary: no learner records in an outside tool. Review gate: I run a bias and accessibility pass before publish. Baseline: 3/3/3/2/3.

Learner checklist

Use this as a final check before submitting. Program leads use a separate review guide when they approve or coach submissions.

  • • Role is named
  • • Workflow goal is specific
  • • Boundary references data, quality, or review
  • • Baseline self-ratings present
Loading your module status…

Next module

AI-assisted work fundamentals →

Keep momentum with the next completion check.