Led · Series A · Jul 2026

Discern Security provides an AI-native security workspace, the Discern Security Loop, that unifies data across an organization's environment and combines reusable AI Skills, intelligent agents, and human-approved workflows to identify, prioritize and remediate control gaps. The platform connects findings to underlying controls, assets and compliance requirements and produces executive-ready reporting aligned with frameworks such as MITRE ATT&CK. Discern says customers typically use less than 25% of features in purchased security products and that its agentic loops can at least double tool utilization. The company plans to expand its library of AI Skills and autonomous capabilities and grow engineering and product teams following its Series A. Discern is based in Sunnyvale, Calif., and raised $13 million in the Series A announced July 30, 2026.

Participated · Series B · Jul 2026

DataBahn provides an agentic data control plane that continuously reduces, enriches, governs and activates telemetry from 600+ sources with neutrality across sources, destinations and AI models. The platform is designed for enterprise scale and security, serving Fortune 500 organizations across industries such as healthcare, financial services, manufacturing and transportation. DataBahn has grown primarily through channel and strategic partners and reports strong commercial metrics, including over 400% year-over-year revenue growth, 180% net revenue retention, zero customer churn, and a 97% proof-of-concept win rate. Headquartered in Dallas, the company emphasizes cost reduction for cloud data movement, storage, compute and AI inference by routing only the data required for real-time operations. Following its Series B, DataBahn plans to invest in R&D and new product capabilities and will preview new functionality at events such as Black Hat USA 2026.

Led · Series C · Jul 2026

Founded in 2019 in San Sebastián, Multiverse Computing builds AI solutions inspired by quantum computing techniques—notably tensor-network approaches—to compress large language models and produce energy-efficient AI. The company says its technology can reduce model size and energy use by up to 95%, deliver 4x–12x faster inference, and cut costs by 50%–80%, enabling deployment on constrained devices like smartphones, laptops and satellites. Multiverse has launched a new generation of optimized models, including an open-source model and a family of compressed models designed for resource-limited environments. It has formed strategic partnerships with firms such as Cohere, Axelera AI, Plain Concepts, Inetum, EY and PwC to accelerate adoption in finance, energy, logistics, space, health and defense. Financially, Multiverse closed a €189M Series B in June 2025 (led by Bullhound Capital and including SETT and other investors) and is currently finalizing a larger round that could total €500M, backed in part by a €107M SETT commitment.

Led · Seed · Apr 2026

Founded in 2025 and headquartered in Tel Aviv, Capsule Security is developing a real-time "runtime trust layer" to provide visibility and control over autonomous AI agents that access data, execute workflows and interact with business systems. The platform is designed to detect and block unsafe or unauthorized agent behavior without requiring proxies, gateways, SDKs or browser extensions, and the company says it integrates with several agent and enterprise platforms including Microsoft Copilot Studio and Salesforce Agentforce. Capsule launched from stealth alongside the publication of two enterprise AI vulnerabilities (ShareLeak and PipeLeak) and an open-source checkpoint tool, ClawGuard. The company is led by CEO Naor Paz and CTO Lidan Hazout. Financially, Capsule announced a $7 million Seed raise led by Lama Partners and Forgepoint Capital International; the articles do not disclose revenue or other operating metrics.

Led · Equity · Mar 2026

Qevlar AI provides an agentic AI SOC platform that autonomously enriches, correlates, and investigates security signals, enabling security teams to shift from reactive alert triage to proactive posture improvement. The software automates data enrichment, pattern recognition, and reporting, giving analysts the time to focus on higher-value tasks such as threat hunting and remediation. Customers including Mercedes-Benz, Sodexo, Orange Cyberdefense, ECI, and Atos report a 10× reduction in investigation time—down to three minutes—while achieving 100 % alert coverage and round-the-clock investigations with no quality loss. With the new capital, the company plans to turn the investigative output into actionable insights that identify root causes and reduce future alert volumes. Founded in 2023 and operating out of Paris and New York, Qevlar AI has quickly built a global footprint among Fortune 500 enterprises and leading MSSPs. The latest $30 million raise underscores strong investor confidence and will fund the evolution from automated investigations to a full end-to-end AI SOC insights platform.