https://store-images.s-microsoft.com/image/apps.29172.a5f2f8cb-3a7a-4826-95c2-7d90a1e5e05e.d6493b3d-87b9-4d0c-9696-bc4769abceba.4c46bb93-f980-4c73-b2f0-5bc309fe3924
DeepTempo
DeepTempo AI, USA
DeepTempo
DeepTempo AI, USA
DeepTempo
DeepTempo AI, USA
Deep-Learning based incident detection
Deep-Learning based incident detection
Core Technology
Tempo is a cybersecurity solution built on a Deep-Learning LogLM (Log Language Model) - the world's first purpose-built foundation model specifically designed for log analysis and cybersecurity threat detection.
The model understands the language of logs, leading to vast improvements in threat detection, isolation, and remediation. Like LLMs, LogLMs are Foundation Models that apply their understanding across very different environments and in response to differing inputs. The difference is Tempo is a transformer based encoder model that is domain specific and is pre-trained using enormous quantities of logs.
KEY CAPABILITIES & FEATURES:
- Anomaly Detection with Exceptional Accuracy: Tempo has demonstrated a unique blend of accuracy and practicality, with false positive and false negative rates lower than one percent after adaptation to a new user's domain.
- MITRE ATT&CK Integration: Tempo automatically tags all stored sequences with the closest MITRE ATT&CK techniques. Using only flow logs, the model can identify whether reconnaissance, lateral movement, data exfiltration, or other common attacks are occurring.
- Agent-less Architecture: Operates with an agent-free architecture. This incident detection is done without the need to send the raw data to SIEM (Security Information and Event Management) systems.
- Foundation Model Approach: Foundation models, like Tempo LogLM, can apply what they learn in one environment to each new environment they encounter. They are collective intelligence - at work in defense of us all.
- Data Efficiency: Tempo also embeds this and other information in compact representations called embeddings, which are less than 1 percent the size of the original logs, enabling faster and more efficient analytics while reducing spending on log storage and analysis.
- Scalability: The LogLM software can run on-premises in any Kubernetes-based workload management system, or in a data lake, and can scale to handle petabytes of log data.
- Continuous Adaptation: Models take in newly encountered data and grow stronger the longer they run. The more models are deployed in the world the more they learn via collective defense and self supervised training.
https://store-images.s-microsoft.com/image/apps.25416.a5f2f8cb-3a7a-4826-95c2-7d90a1e5e05e.d6493b3d-87b9-4d0c-9696-bc4769abceba.5a4ffd52-04e2-46cd-8f07-9ea8c96e45bd
https://store-images.s-microsoft.com/image/apps.25416.a5f2f8cb-3a7a-4826-95c2-7d90a1e5e05e.d6493b3d-87b9-4d0c-9696-bc4769abceba.5a4ffd52-04e2-46cd-8f07-9ea8c96e45bd