Practice Objective
Apply the AI capabilities you've learned so far to your work environment. Follow these three steps: first, identify repetitive tasks; second, connect them with potential technical capabilities; then, select one opportunity for initial analysis.
This is an exploratory exercise with no single correct answer. What's important is to record observations based on actual circumstances, without first proving that a particular AI approach must be effective.
Step 1: Identify repetitive tasks
List three tasks that you or your team frequently perform, which have similar patterns. For each task, record: the amount of time spent per week; the consistency of the process, rated as high, medium, or low. Are there any available data? .
Project record | Task 1 | Task Two | Task Three |
|---|---|---|---|
Task name | To be filled | To be filled | To be filled |
Time spent per week | To be filled | To be filled | To be filled |
Consistency: high, medium, or low | To be filled | To be filled | To be filled |
Are there available data: yes or no | To be filled | To be filled | To be filled |
Step 2: Match the problem with possible abilities
Find a example for each task in the first step. Possible helpful AI capabilities .
Business challenges | Explorable AI capabilities |
|---|---|
A large volume of customer inquiries | Natural Language Processing |
Manually process documents | Computer vision and text extraction |
Decision inconsistency | machine learning models |
Difficulty in resource allocation | Predictive analysis |
These correspondences are for gaining insights and do not yet prove that the approach suits your task. You can first list the possible capabilities that might be used for each of the three tasks.
Step 3: Analyze a specific opportunity
Pick one from the previous challenge and answer four questions:
- What is the current process? write How to complete this task currently .
- What AI methods might be helpful? Corresponding to the ability recognized earlier.
- What benefits might this bring? write Desired improvement result .
- What are the available data sources? Record the actual obtainable, relevant materials suitable for this task.
The four answers can be placed on the same page as a starting point for subsequent discussion and verification.
Note: The worksheet has been organized according to the three tasks and four analysis requirements of the original exercise, for personal practice use. The submit button, scoring, deadline, number of discussions, and personal replies on the original page are not publicly available; this page does not collect or upload any work you complete.
Related Reading
After selecting a worthwhile opportunity for verification, use Comprehensive Operation Opportunity Solution Further supplement the current process, evaluation criteria, impacts, and risks.
Reference and Extended Reading
The following are editorial additions for queries and learning entry points, which you can refer to as needed.
- Google: Introduction to Machine Learning Problem Formulation （Chinese）: Begin with real-world problems to determine if machine learning is appropriate, then define project direction.
- Google: Define model objectives, outputs, and success metrics (Chinese): align expected business outcomes, model outputs, and evaluation criteria.
- Google: Dataset Characteristics （Chinese）: Check data volume, labeling errors, noise, and the relationship between data and model performance.