Warehouse-native entity resolution
Entity resolution that runs as jobs inside your own data platform, so records are matched where they live and IDs are written back to your tables: Zingg on Spark or as a Snowflake Native App (paid editions), and the Splink library on Spark or DuckDB.
Zingg
Warehouse-native entity resolution that matches customer, supplier and product records with ML in your own data platform (Databricks, Fabric, AWS, GCP or Snowflake); open source, with paid editions for persistent IDs.
Best for data engineers who want ML entity resolution as a pipeline step inside their own data platform.
Splink
Open-source probabilistic record linkage library under MIT where a Fellegi-Sunter model estimated without labels scores pairs on DuckDB or Spark, with waterfall charts that explain each match.
Best for teams that want free, explainable entity resolution in Python without labeling data.
Entity resolution engines
Entity resolution engines that load records into their own store or service: an SDK with its own repository, a managed AWS service, or a real-time API that resolves records as they arrive.
Senzing
Entity resolution SDK embedded in your application that resolves each record against existing entities in real time with principle-based rules and explains why records matched.
Best for engineering teams that need real-time entity resolution inside their own infrastructure, including air-gapped sites.
AWS Entity Resolution
Managed entity resolution on AWS that matches S3 records with rule-based, ML or data provider workflows, assigns a match ID, and runs incremental or near real-time rule-based matching.
Best for AWS teams that want pay-per-record entity resolution on S3 data.
Tilores
Real-time entity resolution engine that matches records as they arrive with explicit rules and ML scoring, keeps an auditable entity graph, and serves entities through GraphQL in about 150 ms.
Best for operational entity resolution where every link must be explainable.
Quantexa
Decision intelligence platform whose ML entity resolution resolves entities in batch or at request time and feeds network graphs for financial crime and risk.
Best for banks that want entity resolution tied to graph analytics and investigations.
MDM hubs with built-in matching
MDM hubs whose matching is a product of its own: they resolve entities and then build golden records that data stewards review.
Tamr
MDM platform built on ML entity resolution, where models score record pairs with labeled confidence and low-confidence pairs go to curator queues.
Best for enterprises that want ML entity resolution with steward review inside an MDM platform.
Reltio by SAP
SaaS MDM whose entity resolution combines pretrained LLM matching models with fuzzy and exact match rules, resolving records in real time with persistent Reltio IDs.
Best for teams that need real-time entity resolution feeding operational golden records.
LakeFusion
Databricks-native MDM whose entity resolution runs deterministic rules, then a Random Forest trained on steward decisions, then an LLM for uncertain pairs, with SHAP explanations.
Best for Databricks teams that want explainable entity resolution inside Unity Catalog.
Verato
Healthcare MDM whose referential matching compares patient and provider records with a reference database of 300M+ identities instead of only with each other.
Best for healthcare teams resolving patient and member identities across systems.
IBM Match 360
MDM service whose matching engine combines probabilistic, deterministic and rule-based matching with ML-guided weights and thresholds.
Best for IBM Cloud Pak for Data users that need entity resolution inside an MDM service.
Customer data platforms with identity resolution
Customer data platforms that resolve customer identities for marketing: deterministic rules on emails and IDs, sometimes ML, for people and households first. Listed for comparison; they have no row in the MDM feature matrix.
Hightouch Identity Resolution
Identity resolution that runs in your own warehouse: exact matches on emails, phones and device IDs, then AI probabilistic matching, with profiles for people, households, accounts or any custom entity.
Best for marketing and data teams that want customer profiles built in the warehouse and synced to business tools.
RudderStack Profiles
Identity stitching in your own data cloud: an id_stitcher model links identifiers seen together in your data into an identity graph with one canonical ID per user, household, account or device.
Best for teams already collecting events with RudderStack that need deterministic identity stitching, not fuzzy matching.
Amperity
CDP whose Stitch identity resolution uses machine learning to match customer records from POS, ecommerce, loyalty and CRM systems into profiles with a unique ID, after the data is ingested into Amperity.
Best for consumer brands that want ML identity resolution and customer activation from one vendor.
Twilio Segment Unify
Real-time identity graph that merges each customer's web, mobile, server and partner events into one profile with a persistent ID, using ID rules on cookies, devices, emails and external IDs.
Best for teams already sending events through Segment, on the Business plan.
Evaluating data tools in 2026?
I curate the most comprehensive lists of data tools, up-to-date comparison tables, and actionable market guides for data teams. More useful than a Gartner report. More reliable than asking AI. Next up: Data FinOps.
By Ari Bajo - data engineer turned writer.
Compare master data management tools
The comparison tables behind this list.
How I check every fact
I check each feature, integration and price against the vendor's docs, pricing page or GitHub repository, and link the fact to that source.
Vendors and users can send corrections. I verify them against the docs before changing anything, and every page shows when it was last updated.
Spot a mistake? Send a correction →