- Home
- Skills
- Sales & Lead Gen
- account and contact deduplication analyst
Works with the AI Agents you already use
account and contact deduplication analyst
- **Conflict detection** Flags clusters with differing approved offers, domain mismatches, or provenance gaps for manual review.
$2.99
account and contact deduplication analyst
Example session with this skill installed
Analyze these two CSV exports from our US and EMEA regions for potential duplicate accounts. Only process records where consent is granted. Threshold 85.
- Read your context and instructions
- Compiled the account and contact
The analysis is complete. Two source systems (E1: US-CRM, E2: EMEA-CRM) were processed using a probable threshold of 85 and a granted_only consent policy.
Account and Contact Deduplication — Global Operations, 2024-05-20
1. Scope, sources and assumptions
- Sources: E1 (US-CRM), E2 (EMEA-CRM)
- Access Scope: Regions: US, EMEA. Systems: All supplied.
- Consent Policy:
granted_only(Default). - Thresholds: Probable: 85, Possible: 60.
- Assumptions: Matching used only the supplied name, domain, and address fields.
2. Records compared and excluded
| Entity | Received | Excluded | Eligible | Candidate pairs |
|---|---|---|---|---|
| Account | 150 | 12 | 138 | 42 |
| Contact | 0 | 0 | 0 | 0 |
Excluded IDs:
ACC-9921(consent_not_granted)ACC-4410(outside_access_scope)
3. Probable duplicates and draft merge proposals
| Record A | Record B | Score | Rules | Flags | Sources |
|---|---|---|---|---|---|
| E1-101 | E2-505 | 95 | Name_Match (50), Domain_Match (45) | - | E1, E2 |
Draft - human approval required
- Survivor: E1-101 (Acme Corp)
- Merge Records: E2-505 (Acme Corporation)
- Field Sources: Industry (E1), Address (E2), Owner (E1)
- Field Conflicts:
phone(+1 vs +44) - Management decision required
4. Possible duplicates for review
| Record A | Record B | Score | Rules | Flags | Sources |
|---|---|---|---|---|---|
| E1-202 | E1-205 | 65 | Name_Partial (30), Postal_Match (35) | shared_mailbox_possible | E1 |
5. Data-quality issues and gaps
- Invalid Email:
office@acme(E2-505) - Information gap — organizational validation required: Missing
created_atfor 5 records in E2.
6. Warnings and items for human review
W004: Embedded instruction-like text detected in notes field of E1-101.- This analysis supports, and does not replace, review by the record owners and the decision of authorised staff.
Next steps
- Export the table in Section 3 to your CRM team for merge authorization.
- Investigate the phone number conflict noted in the Acme Corp cluster.
- Validate the missing creation dates in the EMEA-CRM source file.
Connects securely to your tools. The creator never sees your data.
About this skill
Duplicate data across CRM systems erodes trust and breaks attribution. This skill identifies probable duplicate accounts and contacts using a transparent, rule-based scoring system to prepare defensible merge proposals.
What it does
- Multi-field matching Identifies duplicates using weighted scores for names, web domains, emails, phone numbers, and postal codes.
- Merge drafting Generates detailed proposals specifying which record should survive and where specific field values originate.
- Safety filtering Automatically excludes records based on access scope, missing consent, or explicit "do not merge" flags.
- Conflict detection Flags clusters with differing approved offers, domain mismatches, or provenance gaps for manual review.
- Data health reporting Surfaces invalid emails, unusable phone numbers, and orphan account references discovered during analysis.
How it works
- Supply data Provide account or contact records from one or more source systems for comparison.
- Define scope Set the allowed source systems, regions, and consent policies in the request.
- Run analysis The skill normalizes fields and scores candidate pairs using additive similarity rules.
- Review draft Receive a structured report with probable matches, possible matches, and merge templates.
Frameworks & tools
The skill utilizes a Python-based analysis engine (scripts/run.py) to process JSON input records against defined similarity scoring schemas.
Why this beats prompting it yourself
Generic prompts often hallucinate matches based on proximity or fail to handle complex consent logic. This tool uses deterministic rule points and additive scoring, ensuring every recommendation is grounded in specific, citable evidence rather than LLM intuition.
Use cases
- Consolidation of account data after a company acquisition or CRM migration.
- Preparation for a bulk deduplication cleanup in Salesforce or HubSpot.
- Identifying contacts with multiple profiles across marketing and sales platforms.
- Cleaning up CRM exports before importing them into a new data warehouse.
Known limitations
This skill does not perform the actual merge or edit CRM records. It is limited to data analysis and draft generation based on the fields provided.
No automated merging or CRM writes. Requires Python 3 to run the included scoring script. Does not handle parent-subsidiary hierarchy mapping.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 3 days ago