thatDot offers a streaming graph analytics engine that processes real-time and historical data to identify relationships and detect anomalies without time window limitations. This technology enables rapid threat detection and fraud prevention, significantly reducing response times and minimizing financial losses for enterprises.
Funding
Funding not disclosed
CFFounders
Product
Problem
Traditional event stream processing engines are limited by time window constraints, hindering the ability to detect complex, multi-stage threats and anomalies that unfold over extended periods. Analyzing relationships between data points across multiple streams and historical data is computationally expensive, leading to delays in threat detection and increased exposure windows.
Solution
thatDot offers a streaming graph analytics engine that performs real-time, deep analysis of streaming and historical data to identify relationships and detect anomalies without time window limitations. By creating a dynamic graph of data streams, thatDot enables rapid threat detection and fraud prevention, significantly reducing response times and minimizing financial losses. The platform's novelty detection capabilities leverage a proprietary AI to identify unknown threats and reduce false positives by learning the contextual fingerprint of the data environment.
Target Audience
The primary customers are cybersecurity, financial services, and other industries that require real-time analysis of large datasets for threat detection, fraud prevention, and anomaly detection.
Features
- Real-time analysis of structured and unstructured data from multiple streaming sources (e.g., Apache Kafka, Kinesis, SQS) and batch files
- Streaming graph database identifies relationships between data points without time window limitations
- Novelty detection uses proprietary AI to identify unknown threats and reduce false positives
- Cypher graph query language and rich APIs for embedding into applications and data pipelines
- Scalable architecture distributes workload across a cluster, handling large datasets with low resource requirements
- Backpressure systems prevent data loss and overwhelm of downstream applications
- Real-time forensics capabilities for investigating and locating new threat signatures