Samba Identity Resolution

**Overview**
Samba Identity Resolution enables privacy-focused data matching directly within Databricks environments using Delta Sharing. Clients can match identifiers such as Hashed Emails (SHA256), Mobile Ad IDs (MAIDs), CTV Device IDs, and IP Addresses to Samba’s proprietary device graph without moving data.

This solution is designed to streamline identity workflows while meeting strict data governance requirements. By operating entirely within the client’s own environment, it allows for complete control, transparency, and compliance.

**Use cases**
+ **Self-Service Identity Matching:** Organizations can independently match their own hashed identifiers against Samba’s identity graph using SQL queries within Databricks. This enables flexible onboarding of new datasets without delays or manual intervention.
+ **Operational Efficiency and Scalability:** The solution removes the need for manual match table creation, reducing internal support costs and enabling faster activation across multiple use cases, from data licensing to media measurement.
+ **Pathway to Advanced Data Licensing:** Matching in Databricks lays the foundation for expanded collaboration, including future access to Samba’s TV content and ad exposure feeds. This progression enables advanced measurement, attribution, and cross-platform analysis as clients scale.
+ **Privacy-Compliant Analytics and Attribution:** Clients can generate audience insights and measure campaign exposure without sharing raw identifiers. Timestamp fields support recency and frequency analysis for more effective audience planning and performance evaluation.

**Product details**
Clients receive access to a structured table designed for scalable identity resolution, accompanied by schema documentation and example SQL templates to support quick integration.

The table includes fields that allow clients to identify the match type, understand identifier usage, and track when identifiers were first and last seen. All inputs must be pre-hashed using SHA-256. Plaintext identifiers are not accepted.

Dataset represented includes:
- *smba_id* (Samba ID)
- *id* (Identifier)
- *match_key_type* (ID Type)
- *first_time* (First Time Seen)
- *last_time* (Last Time Seen)

**Additional Insights**
+ All matching takes place within the client’s Databricks workspace. Data does not leave the environment at any point.
+ Access is managed through Unity Catalog, supporting enterprise-grade security and governance.
+ The product is designed to replace manual identity matching workflows for Databricks-based clients, supporting faster, more scalable data collaboration.
+ Comprehensive documentation and SQL examples are available upon access approval.

For more information or access requests, please contact: [e-mail]

搜集汇总

背景与挑战

背景概述

Samba Identity Resolution是一个隐私优先的数据匹配解决方案，允许客户在Databricks环境中通过Delta Sharing匹配多种加密标识符与Samba的设备图谱，无需移动原始数据。该方案优化了身份识别工作流程，确保数据治理合规性，适用于广告、媒体测量等多种场景。