AI-native cybersecurity platform for nation-state and critical infrastructure defense
Dream builds an AI-first cybersecurity platform centered on a proprietary Cyber Language Model designed to detect and respond to threats in real time. The tech stack is heavily LLM-oriented (PyTorch, vLLM, TensorRT-LLM, Ray, DeepSpeed) with active adoption of LangChain and LangGraph, signaling a shift toward agentic architectures for threat response. Engineering dominance (21 headcount) paired with dedicated security and research teams reflects a company betting on AI-native detection and response rather than traditional rule-based security.
Notable leadership hires: Cyber Research Team Lead
Dream is an Israeli cybersecurity startup founded in 2022, headquartered in Tel Aviv, focused on defending nations and critical infrastructure from sophisticated cyber and generative AI threats. The platform automates threat discovery, risk calculation, real-time detection, and response planning using AI. Core operational areas span LLM infrastructure, AI detection models, threat intelligence platforms, and CI/CD automation. Active hiring spans engineering, security research, and data roles across Israel, Austria, and Germany, with seniority weighted toward senior and lead engineers.
PyTorch, TensorFlow, vLLM, TensorRT-LLM, Triton Inference Server, Ray, DeepSpeed for model serving and training. Active adoption of LangChain and LangGraph for agent orchestration.
Next-gen cyber threat intelligence platform, AI-driven agent development, AI detection models, LLM serving infrastructure, CI/CD pipelines, and design systems. Key challenge areas include low-latency AI training and scaling detection reliability.
Other companies in the same industry, closest in size