First Pilot Path

For a first pilot, Prodia works best when the goal is not “cover everything”, but to validate one high-value closed loop:

question → analysis → diagnosis → recommendation

1. What the pilot should prove

A first pilot should verify three things:

1. Whether users are willing to ask real business questions in natural language
2. Whether the system can answer one core scenario consistently
3. Whether the answer is good enough to support the next action or collaboration step

In other words, the goal is not to build every page. The goal is to prove that Prodia can move one typical problem from “seeing results” to “supporting diagnosis”.

2. Recommended four-step path

Step	What to confirm	Typical output	Recommended reading
Step 1 Select the problem	Pick a problem that matters, is frequently reviewed, and has visible value	Pilot topic, scope, target users	Typical Scenarios / Capability Boundary and Best-fit Problems
Step 2 Check the data	Confirm what objects, metrics, events, and context are currently available	Minimum data checklist, current capability boundary	Data Readiness and Activation Conditions / Deployment Modes
Step 3 Run the loop	Design one real path from question to diagnosis to recommendation	Demo questions, analysis results, recommended actions	Feature Overview / Typical Workflows
Step 4 Define acceptance	Clarify what “effective” means for this pilot	Acceptance criteria, review conclusions, next-stage scope	Core Value / FAQ

3. Good first-pilot topics

Best starting topics usually share these characteristics:

• The business team genuinely cares and reviews the issue frequently
• Some usable data foundation already exists
• The scope is concrete enough not to expand into an open-ended project

Typical starting points:

• OEE fluctuation analysis
• Yield anomaly attribution
• Downtime / fault impact analysis
• Bottleneck station identification

4. What not to do in the first pilot

• Try to cover all lines, all roles, and all modules at once
• Assume that “data is connected” automatically means “value is unlocked”
• Limit the pilot to interface demos without validating the decision loop
• Move to phase two without clear acceptance criteria

5. A simple readiness check

The first pilot is usually suitable when these three conditions are met:

1. One high-value business question is already defined
2. At least basic object data and core metrics exist
3. The project team is willing to validate one scenario end to end first