Research Paper

Why It Matters

Insider threats and APTs cause significant security breaches but are hard to detect due to complex, evolving patterns. This approach improves detection accuracy without requiring labeled data, enabling faster, scalable threat identification and reducing risk exposure for organizations.

Potential Customers & Pain Points

• Enterprises – Difficulty detecting insider threats early
• Cybersecurity firms – Need scalable accurate threat detection tools
• Government agencies – Require advanced APT detection without extensive training data
• Managed security service providers – Need interpretable alerts to prioritize responses

Research Paper

Why It Matters

Cloud infrastructures face increasingly sophisticated cyberattacks requiring fast, autonomous defense to minimize damage and operational disruption. This solution reduces false positives and detection delays, enabling scalable, continuous protection that adapts to evolving threats. It transforms cybersecurity workflows by integrating intelligent automation for proactive defense.

Potential Customers & Pain Points

• Cloud service providers – Need real-time accurate intrusion detection
• Enterprises with cloud infrastructure – Require automated threat mitigation to reduce manual response
• Managed security service providers – Demand scalable adaptive defense tools
• Government agencies – Need robust protection against evolving cyber threats.

Research Paper

Why It Matters

Software security depends on quickly identifying vulnerabilities within large codebases, a process that is often slow and resource-intensive. Antares accelerates this by delivering near state-of-the-art accuracy with much smaller models, reducing inference time and cost. This scalability and efficiency can transform security operations by enabling faster, more affordable vulnerability detection at scale.

Potential Customers & Pain Points

• Software security teams – Need faster vulnerability detection
• Cybersecurity firms – Require cost-effective analysis tools
• Large enterprises – Struggle with scaling security audits
• DevOps teams – Need integration-friendly low-latency solutions

Research Paper

Why It Matters

Organizations struggle to select effective AI malware detectors due to inconsistent evaluations and overlooked factors like temporal decay and adversarial attacks. EXE-Bench provides a unified, practical assessment that improves deployment decisions, reduces security risks, and optimizes endpoint performance at scale.

Potential Customers & Pain Points

• Enterprise cybersecurity teams – Need reliable malware detection resilient to evolving threats
• Endpoint security vendors – Need benchmarks to validate and improve AI models
• IT administrators – Need efficient malware detection with low computational overhead

Research Paper

Why It Matters

Cybersecurity teams struggle to detect multi-stage attacks spanning system, network, and browser logs due to lack of comprehensive labeled data. This dataset enables more accurate detection by correlating cross-source events with fine-grained ATT&CK labels, improving threat identification and response. It scales to real-world environments by covering diverse attack techniques and telemetry sources.

Potential Customers & Pain Points

• Enterprise security teams – Need comprehensive multi-source data for accurate attack detection
• Security software vendors – Require labeled datasets to train advanced detection models
• Managed security service providers – Need scalable tools to identify complex multi-stage attacks
• Cyber threat researchers – Lack publicly available datasets with detailed ATT&CK labels.

Research Paper

Why It Matters

Insecure code generation by AI models poses significant risks to software security and developer productivity. SPARK reduces vulnerabilities by improving security awareness in code generation without costly retraining or heavy external data reliance. This approach scales across languages and models, enabling safer software development workflows and reducing security audit burdens.

Potential Customers & Pain Points

• Software developers – Risk of generating vulnerable code
• Enterprises – High cost of security audits and remediation
• AI platform providers – Need to improve model security outputs
• Security teams – Difficulty integrating security knowledge into AI tools

Research Paper

Why It Matters

Face presentation attacks threaten biometric security systems, requiring robust detection methods that work under diverse spoofing and capture conditions. This solution reduces computational overhead by eliminating optical flow at inference, enabling real-time deployment on edge devices. It scales to various applications needing secure, efficient face authentication.

Potential Customers & Pain Points

• Mobile device manufacturers – Need efficient accurate face anti-spoofing
• Security system providers – Require real-time spoof detection
• Financial services – Demand robust biometric fraud prevention
• IoT device makers – Need lightweight models for constrained hardware.

Research Paper

Why It Matters

As AI-generated video content becomes widespread, verifying authenticity is critical to combat misinformation and protect digital trust. DYMAPIA offers fast, precise detection of manipulated videos, enabling scalable media verification and secure content filtering workflows. This supports industries reliant on trustworthy visual data, such as journalism, law enforcement, and social platforms.

Potential Customers & Pain Points

• Media companies – Need reliable fake video detection
• Social media platforms – Need scalable content moderation
• Law enforcement agencies – Need accurate forensic tools
• Misinformation watchdogs – Need early detection of manipulated media

Research Paper

Why It Matters

Backdoor attacks on large language models pose serious security risks, undermining trust and safety in AI applications. TIGS offers a practical defense that preserves model performance and speed, enabling safer deployment of LLMs in sensitive or high-stakes environments. This scalable approach supports broad adoption across industries relying on secure AI.

Potential Customers & Pain Points

• AI platform providers – Need to secure LLMs from backdoor attacks
• Enterprises deploying LLMs – Require reliable low-latency defenses
• Cloud service operators – Must maintain model integrity without costly retraining
• Security-focused AI developers – Seek practical plug-and-play mitigation tools

Research Paper

Why It Matters

IoT devices face increasing cyber threats due to expanded attack surfaces and limited security capabilities. A-THENA enhances early detection accuracy and reduces false positives, enabling timely responses to attacks. Its lightweight design supports deployment on resource-constrained devices, facilitating scalable and practical IoT network protection.

Potential Customers & Pain Points

• IoT device manufacturers – Need robust low-latency security
• Smart home providers – Require early threat detection
• Industrial IoT operators – Need scalable intrusion detection
• Network security firms – Demand improved detection accuracy with low false alarms

Research Paper

Why It Matters

Connected vehicles rely on Controller Area Network communications that are vulnerable to cyberattacks, risking safety and operational integrity. DAIRE offers fast, accurate detection of these attacks with minimal computational overhead, enabling scalable real-time protection in automotive systems. This improves vehicle security and supports safer, more reliable Internet of Vehicles deployments.

Potential Customers & Pain Points

• Automotive manufacturers – Need robust real-time cybersecurity for connected vehicles
• Fleet operators – Require efficient detection of vehicle network attacks to avoid downtime
• IoV platform providers – Demand scalable security solutions with low latency
• Automotive cybersecurity firms – Seek advanced detection models to enhance product offerings.

Research Paper

Why It Matters

Unknown network threats pose significant risks as they evade traditional detection systems, especially in imbalanced data scenarios. RPM-Net improves detection accuracy and interpretability, enabling security teams to identify and respond to emerging threats more effectively. This enhances cybersecurity resilience and reduces potential damage from novel attacks.

Potential Customers & Pain Points

• Enterprises – Difficulty detecting unknown cyber threats
• Security vendors – Need for interpretable threat detection models
• Cloud providers – Managing imbalanced threat data
• Government agencies – Enhancing national cybersecurity defenses

Research Paper

Why It Matters

Cybersecurity faces a shortage of skilled professionals and frequent human errors causing breaches. SentinelSphere addresses these by integrating precise AI threat detection with accessible, real-time security training, improving organizational defenses and user awareness. This dual approach scales across industries, reducing incident rates and operational risks.

Potential Customers & Pain Points

• Enterprises – Need to reduce security breaches caused by human error
• SMBs – Lack resources for advanced threat detection and training
• Educational institutions – Require effective cybersecurity training tools
• Managed security service providers – Need scalable detection and user education solutions.

Research Paper

Why It Matters

Software vulnerabilities pose critical security risks and require timely detection to prevent exploits. VulGNN offers a scalable, efficient alternative to large language models, enabling faster retraining and deployment in development pipelines. This reduces detection costs and accelerates secure software delivery across diverse codebases.

Potential Customers & Pain Points

• Software development companies – Need fast accurate vulnerability detection
• Cybersecurity firms – Require scalable customizable analysis tools
• DevOps teams – Seek lightweight models deployable at the edge
• Enterprises – Demand cost-effective security solutions integrated into CI/CD pipelines

Research Paper

Why It Matters

LLM agent systems often suffer security failures due to vulnerabilities in surrounding software stacks rather than model weights alone. Agent Audit helps developers identify and mitigate these risks early, reducing potential breaches and operational failures. This improves trust and safety in deploying AI-driven agent applications at scale.

Potential Customers & Pain Points

• AI developers – Need to identify hidden security risks in agent code
• DevOps teams – Require automated security checks integrated into CI/CD
• Enterprises deploying LLM agents – Need to prevent credential leaks and privilege escalations.

Research Paper

Why It Matters

Industrial control systems face increasing cyber threats that can cause costly shutdowns and operational disruptions. This solution improves attack detection accuracy and discriminates attack types, enabling continuous safe operation and reducing false alarms and costs. It scales to critical infrastructure, enhancing security and operational resilience in real time.

Potential Customers & Pain Points

• Critical infrastructure operators – Need to prevent costly shutdowns from cyber-attacks
• Industrial control system providers – Need accurate real-time attack detection and classification
• Cybersecurity firms – Need advanced tools for resilient control and anomaly discrimination.

Research Paper

Why It Matters

Generative AI content poses growing risks to content safety and privacy across industries. Existing defenses are fragmented and ineffective against mixed generative models, creating security gaps. A universal, architecture-agnostic defense streamlines protection workflows, reduces vulnerability, and scales to emerging generative threats, enhancing trust and compliance.

Potential Customers & Pain Points

• Social media platforms – Need to detect and block AI-generated harmful content
• Content moderation services – Require scalable defenses against diverse generative attacks
• Enterprises – Need to protect brand integrity from synthetic media
• Cloud security providers – Must secure AI-generated data pipelines
• Government agencies – Need reliable tools to counter misinformation and deepfakes.

Research Paper

Why It Matters

Prompt injection attacks threaten AI system integrity and require efficient, reliable detection at scale. This solution offers a lightweight, deterministic screening layer that reduces latency and dependency on large models, enabling safer AI deployments and scalable security workflows. It improves early threat detection while maintaining auditability and speed.

Potential Customers & Pain Points

• AI platform providers – Need low-latency reliable prompt injection detection
• Enterprises deploying AI assistants – Require scalable auditable security layers
• Cloud service providers – Need to reduce inference costs and improve throughput.

Research Paper

Why It Matters

Phishing causes widespread credential theft and financial fraud, challenging existing detectors that rely on external data and heavy computation. SpecularNet offers a practical, scalable solution with fast inference on standard hardware, improving real-world phishing defense and enabling broader adoption across industries.

Potential Customers & Pain Points

• Cybersecurity firms – Need scalable low-latency phishing detection
• Enterprises – Require cost-effective real-time web threat protection
• Cloud providers – Seek efficient phishing detection without heavy resource use
• Financial institutions – Demand robust defense against evolving phishing attacks

Research Paper

Why It Matters

Targeted attacks on LLMs threaten the trustworthiness of AI systems by covertly forcing malicious outputs. DualSentinel offers a practical, low-cost solution that integrates seamlessly with deployed models, enabling organizations to secure AI services without disrupting normal operations or requiring deep system access. This scalability supports widespread adoption across industries relying on LLM APIs.

Potential Customers & Pain Points

• AI service providers – Need to secure LLM APIs from stealth attacks
• Enterprises deploying LLMs – Require real-time low-cost attack detection
• Cloud platform operators – Must maintain trust and compliance with minimal overhead.