FAQ's AI Map
- Rajeev Rai
Analytics
AI Map Training Guide FAQ
What is AI Map, and how does it help in data mapping?
AI Map is an advanced tool that automates data mapping between source and target systems. It uses AI to suggest mappings with confidence levels (High, Medium, and Low), reducing manual effort and speeding up integration tasks. Its library feature allows users to save and reuse mappings for similar scenarios, further increasing efficiency.
What are the confidence levels, and how do they work?
High Confidence: Generated for exact matches or previously saved mappings from the library. These can often be applied automatically.
Medium Confidence: Suggested for partial matches where fields are somewhat similar, requiring user review.
Low Confidence: Generated when AI detects low similarity or ambiguity; these mappings must be reviewed and approved manually.
What should I do if a mapping is incorrect or missing?
You can:
Modify or adjust incorrect mappings directly in the mapping layout.
Reject or remove unwanted suggestions without affecting other mappings.
For missing mappings, manually create them and save them to the library for future reuse.
How does the AI Map Library work?
The AI Map Library stores manual mappings created by users. These saved mappings are presented as High Confidence suggestions when similar layouts are encountered, allowing for faster and more accurate mapping in future projects.
How can I apply mappings based on confidence levels?
AI Map allows users to filter and apply mappings by confidence level.
Use High Confidence mappings directly for reliable results.
Manually review Medium and Low Confidence mappings to ensure accuracy before applying them.
What are the key differences between Design Time and Runtime in AI Map?
Design Time: Users create, test, and finalize mappings before deploying them. This ensures the mappings and rules are accurate.
Runtime: Mappings are applied to data during live operations. AI-generated suggestions are reviewed by users to prevent errors before the data is passed to target systems.
Does AI Map support handling complex mapping scenarios?
Yes. AI Map can recognize different naming conventions (e.g., "First Name" vs. "F Name") and intelligently map them. Users can also define custom mappings and rules for complex cases and save them in the library for future use.
Are there any planned updates to improve AI Map?
Planned features include:
Support for natural language-based rule definitions (e.g., "convert date formats").
Enhanced AI-driven suggestions for Medium and High Confidence mappings.
Warnings for unmapped mandatory fields during the design phase.
What are the common issues users face with AI Map?
Missing Mappings: AI Map currently doesn’t alert users about unmapped required fields, so manual checks are necessary.
Accuracy of Low Confidence Suggestions: These require thorough review and manual approval.
10. What are the best practices for using AI Map effectively?
Start with Auto Mapping to save time.
Always review low-confidence mappings for accuracy.
Save manual mappings to the library for better efficiency in future projects.
Utilize the design-time review process to test mappings before runtime deployment.
11. Can AI Map handle mismatched field naming conventions?
Yes, AI Map uses AI capabilities to recognize and map fields with equivalent meanings even if the naming conventions differ, such as "Last Name" and "Surname."
12. Can I define custom rules in AI Map?
Currently, custom rules can be implemented manually, but AI Map is evolving to allow rule definitions in plain language, like "truncate values" or "convert date formats," for automating intricate mappings.
13. Why is manual review necessary for some mappings?
Manual review ensures that Medium and Low Confidence mappings meet system requirements and prevent errors in the data integration process. It is crucial for maintaining accuracy and reliability.
14. How do I handle unmapped mandatory fields?
Perform a manual check during the design phase to ensure all required fields are mapped. A feature for automatic alerts on unmapped mandatory fields is planned for future releases.