Senzing Launches Agentic Entity Resolution for Apache Spark

New Spark-native offering makes Senzing the first entity resolution vendor to offer batch, transactional, and hybrid deployment models with full end-to-end agentic automation

LAS VEGAS--(BUSINESS WIRE)--Senzing, an identity intelligence company, today announced the opening of its Senzing for Apache Spark beta program, bringing the company’s industry-leading entity resolution technology to distributed batch workloads for the first time. Organizations running Spark on AWS EMR, Databricks, or Snowflake can now resolve and relate billions of records across multiple data sources—from fully autonomous data profiling and preparation through to publishing the resolved entity graph to downstream systems.

The launch marks a significant milestone for the entity resolution market. Until now, enterprises faced a binary choice: batch processing systems built on Spark, or real-time transactional systems. Senzing for Spark eliminates that tradeoff. With this release, Senzing becomes the only entity resolution vendor to offer all three entity resolution deployment modes—Spark batch, transactional SQL, and hybrid.

“Picking an entity resolution vendor has long forced a binary choice: Batch Spark or Transactional SQL. We’re excited to turn this ‘or’ into an ‘and.’ With Senzing for Spark, customers get all the intelligence found in our real-time SDK—principle-based entity resolution, entity-centric learning, relationship awareness, global name, address and cross-script matching, and explainability—running natively inside their Spark platform of choice.
— Brian Macy, Head of Operations and Engineering, Senzing

Fully Agentic from Preparation to Publication

Senzing® entity resolution for Spark is designed for agentic AI workflows end-to-end. Powered by the Senzing MCP Server, AI agents execute each stage of the pipeline autonomously:

• Data preparation and mapping: Agents profile, prepare, map, and validate each data source to Senzing-ready dataframes autonomously.
• Distributed entity resolution: With validated dataframes, agents trigger and manage distributed entity resolution jobs across the Spark cluster, executing across all data sources in parallel at any scale.
• Publishing the resolved entity graph: Agents propagate results to any downstream destination—Elasticsearch, knowledge graphs, data lakes—or implant the resolved entity graph directly into an existing live Senzing instance, giving real-time systems an immediate entity intelligence boost.

Availability and Roadmap

Senzing for Spark v1.0, entering beta testing with select partners, supports multi-source batch entity resolution on AWS EMR, Databricks, Snowflake, and standalone Apache Spark deployments. The resolved entity graph output can also be used to pre-populate a Senzing real-time SQL instance.

Senzing for Spark v2.0 (Hybrid), next on the roadmap, will allow organizations to splice batch entity resolution results directly into a live transactional Senzing instance with no downtime and no record-by-record ingestion, enabling rapid onboarding of large new datasets at Spark speed.

Organizations with a Spark cluster and active use cases in financial crime detection, insurance fraud, national security, or customer 360 are encouraged to apply for the beta program.

For more information or to apply for early access, visit Senzing Agentic Entity Resolution for Apache Spark.

About Senzing

Senzing delivers the identity intelligence organizations need to achieve their agentic AI aspirations. As the creator of Agentic Entity Resolution, Senzing enables AI agents to autonomously identify and act on real-world entities in real time or batch—keeping all data secure within customer infrastructure. Backed by 40+ years of innovation and 300+ years of combined team experience, Senzing is trusted by organizations worldwide to ensure their AI agents operate on accurate and trustworthy data. Senzing is headquartered in Las Vegas, Nevada. For more information, visit www.senzing.ai.