AI-based Master Recod Decision
AI-Powered Master Record Selection enhances the DeDupeD merge experience by intelligently identifying the most suitable master record during duplicate record consolidation in Microsoft Dynamics 365.
Instead of requiring users to manually evaluate multiple duplicate records and determine which record should survive the merge, DeDupeD leverages artificial intelligence to analyze all duplicate records and recommend the optimal master record based on data completeness, consistency, reliability, and business relevance.
The AI recommendation includes:
Suggested Master Record
Confidence Score
Selection Reasoning
Complete Audit Trail of the AI Decision
Users remain in full control and can override the AI recommendation before completing the merge.
How AI Master Selection Works
When duplicate records are selected for merge, DeDupeD sends the duplicate record data to the configured AI model for analysis.
The AI evaluates:
Record completeness
Address information quality
Contact information availability
Business identifiers
Data consistency
Record reliability
Overall information richness
Based on the evaluation, the AI recommends the most suitable master record and assigns a confidence score.
Step 1 – Select Duplicate Records for Merge
Navigate to the duplicate records identified by DeDupeD.
Select the duplicate records that should be merged and click Merge.
The merge wizard opens, displaying
Current Record
Duplicate Records
Record Details
Merge Actions
What Happens
DeDupeD collects all selected duplicate records and prepares them for AI analysis.
Users can choose whether AI suggestions are loaded automatically when the merge interface opens or generated on demand by enabling the AI Suggestions option during the merge process.

DeDupeD collects all selected duplicate records and prepares them for AI analysis.
Users can choose whether AI suggestions are loaded automatically when the merge interface opens or generated on demand by enabling the AI Suggestions option during the merge process.

Step 2 – AI Analysis Begins
Once the merge screen loads and AI Master Selection is enabled, DeDupeD automatically begins evaluating the duplicate records.
During analysis:
Duplicate records are compared
Data completeness is evaluated
Address information is reviewed
Contact information is analyzed
Business identifiers are examined
Loading indicator displays:
"Getting AI suggestions..."

Note: when AI Suggestion checkbox (on Merge UI) is unchecked then it follows the manual rule and Suggest Master based on the configured rule/s.
Step 3 – AI Recommends the Master Record
After analysis is completed, DeDupeD highlights the recommended master record.
The selected record displays:
AI Suggested (XX%)
The confidence score indicates how strongly the AI believes the record should be selected as the surviving master.
Example:
92% Confidence
85% Confidence
78% Confidence
Higher confidence indicates stronger certainty in the recommendation.

Step 4 – Review the AI Recommendation
The recommended master record is automatically selected.
The following indicators appear:
Suggested Master Badge: Identifies the AI-recommended master record.
Confidence Score: Displays the confidence percentage for the recommendation.
Master Selection Indicator: Shows which record is currently selected as the master.
Disclaimer
A notification appears:
AI-generated suggestions may not always be accurate. Please review before applying.
Users should review the recommendation before proceeding.
Step 5 – View the AI Explanation
Users can click the information icon beside the AI recommendation to view the reasoning behind the selection.
The Why This Suggestion? panel displays:
Confidence Score
Example: 92%
Explanation: A plain-language explanation describing why the record was selected.
Example:
This record contains comprehensive and valid business information, including multiple professional email addresses, a complete address, and a detailed company description, ensuring high reliability.

Step 6 – Modify the Master Record (Optional)
Users can override the AI recommendation at any time.
To select a different master record:
Choose another record.
Click Set Master.
Review the updated preview.
The AI recommendation remains visible for reference, but the final decision always belongs to the user.
Human-in-the-Loop Governance
AI recommendations are advisory only and never automatically complete the merge.
Step 7 – Complete the Merge
After reviewing the AI recommendation:
1. Verify the selected master record.
2. Review field-level values.
3. Click Finish.
DeDupeD:
Merges duplicate records
Retains the selected master record
Deactivates subordinate records
Preserves related data associations

Step 8 – Review Merge Results
After the merge is completed, DeDupeD creates a merge result record.
Navigate to:
DeDupeD → Job Runs
The merge result displays:
Total Records
Duplicate Count
Merge Count
Processing Status

Step 9 – View Master Selection Details
Open the merged record result.
The Master Details section contains the complete AI audit trail.
Information displayed includes:
Selection Method
Example: AI Master Selection
Selected Master Record
The record chosen by AI as the surviving record.
Confidence Score
Example: 85%
Reason
The AI-generated explanation describing why the record was selected.
Subordinate Records
List of records merged into the master record.
This information provides complete transparency and auditability.

AI Master Selection UI Elements Reference
Element
Description
AI Suggested Badge
Indicates the AI-recommended master record.
Confidence Score
Shows the AI confidence level for the recommendation.
Suggested Master Label
Identifies the recommended surviving record.
Why This Suggestion?
Displays AI reasoning and confidence score.
Master Selection Details
Stores audit information for completed merges.
Selection Method
Identifies whether the master was selected manually or by AI.
Important Notes
AI Recommendations Are Advisory
AI recommendations assist users but do not automatically perform merges.
User Override Supported
Users can select a different master record at any time before completing the merge.
Audit Trail Maintained
All AI decisions, confidence scores, and reasoning are stored for future reference.
Confidence Scores
· 90–100% → High Confidence
· 75–89% → Moderate Confidence
· Below 75% → Review Recommended
Processing Time
AI analysis time depends on:
· Number of duplicate records
· Number of populated fields
· AI model response time
Most analyses complete within a few seconds.
For further queries, reach out to us at crm@inogic.com
Last updated
Was this helpful?