Workshop contents

Session 1 · Context Engineering

Context Engineering

Giving AI the business rules a manager already knows

Now we move from answering a question to understanding a business. The AI receives the same data, plus the company’s rules, targets, and constraints. When a fact is not in the dataset, it is labeled as a controlled workshop assumption.

Context Map

Context typePlain meaningHow it changes the answer
BaselineWhat is happening nowThe AI knows the starting point instead of guessing.
TargetWhat the company wants to improveThe recommendation has a clear direction.
ConstraintWhat the company cannot breakThe AI avoids answers that look smart but cannot be approved.
HistoryWhat already worked or failedThe AI does not repeat mistakes the business already learned from.

Add business context

PromptWorkshop ScenarioClick the text, then ⌘C or Ctrl+C
COMPANY CONTEXT:
- We are a business-to-business distributor serving corporate and home office customers
- Service promise: deliver within the scheduled shipping days
- Leadership target: reduce late delivery rate from 54.8% to below 30% in 6 months
- How current orders are shipped: Standard Class 59.7%, Second Class
  19.5%, First Class 15.4%, Same Day 5.4%
- Budget rule: delivery cost cannot increase more than 12%

DATASET: [DataCo supply chain data attached]

Given this context, analyze our late delivery problem and recommend an
action plan in phases that stays within budget. Prioritize by profit impact.

Format the plan into three phases: Immediate (0-30 days), Short-Term
(30-90 days), and Medium-Term. For each phase, explicitly state the
expected budget impact against the 12% ceiling.

The 54.8% starting rate and the shipping mix percentages come from the data Derived Evidence. The service promise, leadership target and budget ceiling are invented for the exercise Workshop Scenario.

Add history, or memory

PromptWorkshop ScenarioClick the text, then ⌘C or Ctrl+C
PREVIOUS ANALYSIS (Q1):
- We shifted 15% of Second Class order lines to Standard Class in West of USA
- Result: late deliveries dropped 8% in West of USA but increased 3% in
  US Center
- Possible reason: the Standard Class delivery provider has limited US Center
  capacity

Given this history, update your recommendations. Do not repeat strategies
that failed. Account for delivery provider capacity differences by region: what each
delivery provider can actually handle in each area.

Provide a "Revised Action" table contrasting the Q1 approach with the new
approach for US Center.

Why This Matters

Key takeaway

Anyone can access the same AI model. Your competitive advantage is the context you give it: your rules, your constraints, your history.

How to read the figures on this page Source Dataread straight from the DataCo dataset Derived Evidencecalculated from those values Workshop Scenarioinvented for the exercise; not in the dataset