We help defence systems survive attacks and adapt in real time — whether we're hardening what you already run or engineering it from the ground up.
Trusted where failure is not an option
Ten years across IT, OT, and defence systems — including security solutions operating in Ukraine's power grid since the systemic attacks of 2016–2017, and through the sustained cyber-physical attacks since 2022; and architecting, building, and securing integrated targeting, situational-awareness, and force-level C2 systems used in active conflict.
We work on defence systems alongside the operators who run them and the vendors who build them, from system design through to field deployment, with dozens of projects spanning Ukrainian and international defence and dual-use manufacturers.
UA DroneIDSecure IFF capability between drones, operators, and command structures, designed and used in active combat. Protects high-value assets and increases the operational efficiency of unmanned systems operating under continuous cyberattacks, EW, and active enemy countermeasures.
Ukraine’s primary C2 system operates under daily nation-state-level cyberattacks. Cossack Labs works with the DELTA team as a security engineering partner: helping build and maintain a product security programme, as well as hardening more than 30 integrated systems so that the ecosystem remains operational, secure, and resilient under continuous adversarial pressure.
MINISTRY OF DEFENCE OF UKRAINEThe programme that let the Ministry of Defence of Ukraine assess and mitigate the risks across its defence information and communication systems. Grounded in first-hand understanding of how adversaries operate against defence systems in active conflict.
What our partners say
The security engineering team cares not only about security but also about the product itself. It creates real alignment, making security feel like a natural part of the process rather than just a compliance checkbox.
Center of Innovations and Defense Technologies Development of the Ministry of Defence of Ukraine
Identify weaknesses across legacy and modern systems — hardware, software, and the interfaces between them — before an adversary does, so your teams keep operational tempo without disruption.
We work alongside defence agencies and public-sector organisations to navigate complex security challenges — supporting decisions and shaping policy, grounded in deep technical understanding.
Bring defence products and emerging capabilities to field deployment with confidence they will hold under adversarial conditions — from architecture review through to live-environment deployment.
When existing standards don't cover the problem, we build what's missing: new protocols, security controls, and methodological frameworks. We assess emerging capabilities for the weaknesses standards haven't yet caught up to.
Cross-domain integration framework for mission-critical systems
Fabric integrates heterogeneous battlespace assets and moves data securely to the systems that command and decide — across untrusted networks and under contested, degraded, or disconnected (DDIL) conditions.
Proven in the operational conditions of Ukraine, where both military capabilities and national infrastructure operate daily under real-world attack.
Build and operate fleets of secure edge hardware
Edge hardware runs outside controlled environments, on firmware that must stay trustworthy, managed by systems that may themselves be under pressure — across heterogeneous hardware, third-party operators, contested supply chains, and technically capable adversaries.
Fleet enables managing that at scale across the whole device lifecycle in a single management layer. Its trust model lets architects balance pragmatic usability against trust. Hardened in active conflict — across UAVs operating in contested, disconnected environments and edge devices on national power-grid infrastructure.
Securing AI from model build to field inference
Deploying AI in defence introduces security problems standard MLOps tooling was never built for: models can be extracted or manipulated, inference queries leak intent, training data carries re-identification risk, and edge accelerators break conventional build pipelines.
Helmet brings together our work on these problems across the AI lifecycle — dataset and training-environment security, security of AI-enabled applications, and protection of models deployed at the edge — used in active conflict, around real threats and operational requirements.
Let's see it.
Get whitepaper
Apply for the position
Our team will review your resume and provide feedback
within 5 business days