Portfolio

Malicious code detection for open source: analyses what packages actually do to catch supply-chain threats that vulnerability databases miss.

AI-native digital forensics and incident response: a forensic-grade LLM that automates investigation, triage, and containment for teams without a 24/7 SOC.

Detects and responds to disinformation campaigns affecting organisations, using quantitative analysis of narratives across channels.

Threat-focused vulnerability management: continuous threat modelling from real findings, root-cause analysis, and context for developer workflows.

Supervision and fine-tuning for AI agents: full run traces, anomaly review, and training signals from production behaviour.

Identity and access governance on autopilot: discover identities and access paths, automate reviews, onboarding, and offboarding.

Threat hunting workbench: federated search across tools, guided investigation, and reusable hunt trails.

Authorised push payment (APP) fraud intelligence: surface mule accounts and scam-related payment paths before funds move.

FPGA-accelerated, low-latency network encryption for data in motion, with post-quantum crypto agility and high throughput.

AI security testing and runtime protection: discover, assess, and defend models, agents, and AI applications.

CREST-approved, AI-augmented penetration testing with continuous scanning and threat intelligence in one subscription platform.

AI-assisted binary analysis for reverse engineering: models and APIs for working with machine code and assembly.

Security stack intelligence: map tools to frameworks (e.g. MITRE, NIST), measure adoption, and align spend with coverage and renewals.

Automated deception technology: deploy and maintain decoys (seeds) in cloud environments for high-fidelity intrusion detection.