Account Deduplication
“For each account in the CRM, which other accounts represent the same company?”
Book a demo and get a free trial of the full platform: data science agent, fine-tune capabilities, and forward-deployed engineer support.
By submitting, you accept the Terms and Privacy Policy.
Loved by data scientists, ML engineers & CXOs at
A real-world example
For each account in the CRM, which other accounts represent the same company?
B2B CRMs average 20-30% duplicate accounts — "Acme Corp", "ACME Corporation", "Acme Inc." all exist as separate records. Sales reps unknowingly compete for the same account. Revenue attribution breaks. Kumo detects duplicates through shared contacts, overlapping opportunity patterns, and domain relationships. Each duplicate account costs $1,000-5,000 in wasted sales effort and misallocated pipeline value.
How KumoRFM solves this
Relational intelligence for identity resolution
Kumo connects accounts to their contacts, opportunities, email domains, and interaction histories in a unified relational graph. Instead of fuzzy-matching company names, Kumo learns that Account A-101 and Account A-287 share 3 contacts, have overlapping opportunity timelines, and their email domains resolve to the same parent organization. The link prediction model identifies which accounts represent the same company — even when names, addresses, and domains differ across subsidiaries and acquisitions.
From data to predictions
See the full pipeline in action
Connect your tables, write a PQL query, and get predictions with built-in explainability — all in minutes, not months.
Your data
The relational tables Kumo learns from
ACCOUNTS
| account_id | company_name | domain | industry | region |
|---|---|---|---|---|
| A-101 | Acme Corp | acme.com | Technology | West |
| A-287 | ACME Corporation | acme-corp.com | Tech | West |
| A-450 | GlobalTech Inc | globaltech.io | Software | East |
CONTACTS
| contact_id | account_id | title | phone | |
|---|---|---|---|---|
| CON-001 | A-101 | sarah@acme.com | VP Sales | 555-1001 |
| CON-002 | A-287 | sarah@acme-corp.com | VP of Sales | 555-1001 |
| CON-003 | A-450 | mike@globaltech.io | CTO | 555-2050 |
OPPORTUNITIES
| opp_id | account_id | amount | stage | timestamp |
|---|---|---|---|---|
| OPP-201 | A-101 | $250,000 | Negotiation | 2025-09-10 |
| OPP-202 | A-287 | $250,000 | Proposal | 2025-09-12 |
| OPP-203 | A-450 | $180,000 | Discovery | 2025-09-14 |
Write your PQL query
Describe what to predict in 2–3 lines — Kumo handles the rest
PREDICT LIST_DISTINCT(CONTACTS.ACCOUNT_ID, 0, 30, days) FOR EACH ACCOUNTS.ACCOUNT_ID
Prediction output
Every entity gets a score, updated continuously
| ACCOUNT_ID | MATCHED_ACCOUNT_ID | SCORE | TIMESTAMP |
|---|---|---|---|
| A-101 | A-287 | 0.97 | 2025-10-01 |
| A-450 | A-612 | 0.81 | 2025-10-01 |
| A-330 | A-775 | 0.73 | 2025-10-01 |
Understand why
Every prediction includes feature attributions — no black boxes
Account A-101 (Acme Corp)
Predicted: 97% match with A-287 (ACME Corporation)
Top contributing features
Shared contacts (same phone/name)
3 contacts
33% attribution
Opportunity amount overlap
$250K match
25% attribution
Domain parent relationship
acme.*
20% attribution
Industry + region match
Tech / West
13% attribution
Contact email domain similarity
0.92
9% attribution
Feature attributions are computed automatically for every prediction. No separate tooling required. Learn more about Kumo explainability
PQL Documentation
Learn the Predictive Query Language — SQL-like syntax for defining any prediction task in 2–3 lines.
Python SDK
Integrate Kumo predictions into your pipelines. Train, evaluate, and deploy models programmatically.
Explainability Docs
Understand feature attributions, model evaluation metrics, and how to build trust with stakeholders.
Bottom line: Merge 20-30% duplicate B2B accounts — eliminating internal competition, fixing pipeline attribution, and recovering $1-5K in wasted sales effort per duplicate.
Related use cases
Explore more entity resolution use cases
Topics covered
One Platform. One Model. Predict Instantly.
KumoRFM
Relational Foundation Model
Turn structured relational data into predictions in seconds. KumoRFM delivers zero-shot predictions that rival months of traditional data science. No training, feature engineering, or infrastructure required. Just connect your data and start predicting.
For critical use cases, fine-tune KumoRFM on your data using the Kumo platform and Data Science Agent for 30%+ higher accuracy than traditional models.
Book a demo and get a free trial of the full platform: data science agent, fine-tune capabilities, and forward-deployed engineer support.
