Fractal Diamond Refinement Process
FDRP is our proprietary recursive methodology for building complex systems. Each phase is self-similar, containing quality gates, dual-model verification, and continuous improvement loops at every scale.
Seed
Origin capture and problem crystallization. Every project begins with immutable requirements and measurable objectives.
The Seed phase captures the original brief, establishes project boundaries, and creates the origin document that anchors all subsequent work. Quality gate: origin traceability verified, scope boundaries explicit, success metrics quantified.
Refine
Dual-model analysis with aviation-grade verification. Independent verification layers catch what single-pass review misses.
Every critical decision passes through independent verification by multiple AI models and human experts. The Swiss Cheese model ensures no single failure mode can propagate. Quality gate: dual verification receipts for all critical paths.
Converge
Fractal compression of solutions into deployable artifacts. From concept to production-grade output.
Solutions are iteratively compressed through presenter-critique loops until they reach production quality. Each iteration applies Ishikawa fishbone analysis to identify root causes, not symptoms. Quality gate: all computed values derived from measured data (never estimated).
Harden
Production hardening with continuous improvement loops. Systems that monitor themselves and evolve under operational stress.
Deployed systems include built-in Kaizen loops: error detection drives immediate correction, repeated patterns escalate automatically, and effectiveness is tracked over time. Quality gate: self-monitoring active, error budget within tolerance, rollback capability verified.
Where cognition meets criticality
We operate exclusively in domains where system failure has catastrophic consequences. Our cognitive systems are built to the standards these environments demand.
Defence & Security
NATO-aligned cognitive ISR, autonomous threat assessment, and C2 decision support systems. Built for interoperability with allied force structures.
Energy Infrastructure
Predictive maintenance for critical energy systems, digital twins for power generation, and real-time anomaly detection for grid resilience.
Industrial AI
Process optimization and autonomous quality control for manufacturing environments. Self-learning systems that improve under operational pressure.
INDIGO AirGuard
Autonomous air quality monitoring with predictive analytics for urban and industrial environments. INDIGO AirGuard demonstrates our full cognitive stack: edge sensing, real-time analytics, predictive modelling, and autonomous response.
Developed for consortium partnerships across European urban infrastructure projects. The system combines FPGA-accelerated edge processing with cloud-based digital twins to provide sub-second anomaly detection and predictive alerts for air quality degradation events.
EU defence & innovation pipeline
Fraktiq is actively positioning within the European defence and innovation ecosystem, targeting high-impact programmes that align with our cognitive systems capabilities.
DIANA
Defence Innovation Accelerator for the North Atlantic
NATO's premier deep-tech accelerator. Targeting cognitive ISR and autonomous monitoring challenge topics.
2027 ApplicationEDF
European Defence Fund
Consortium-based R&D funding for cross-border defence technology development. Multi-nation partnership forming.
2026 CallAGILE
NATO Innovation Programme
Rapid prototyping and validation of defence technology solutions within NATO frameworks.
2026-2027EDIP / BraveTechEU
European Defence Industrial Programme
Strengthening EU defence industrial base. Supporting SME integration into strategic supply chains.
2026 MonitoringFunding Pipeline Timeline
Strategic milestones from company formation to first consortium deliverables.
European partner network
Multi-nation consortium spanning EU and NATO member states, combining complementary capabilities across the defence and industrial landscape.
Romania
HQ. R&D, cognitive systems, FPGA, methodology IP
Poland
Urban infrastructure, manufacturing (Buglo, TechRamps)
Germany
Urban design, playground equipment (Playparc)
Netherlands
Street furniture, urban infrastructure (HeBlad)
Estonia
Cyber defence ecosystem, digital governance
Finland
Defence technology corridor, Arctic capabilities
Engineering stack
Purpose-built for environments where deterministic correctness outweighs development velocity. We compute, not estimate.
Romanian R&D fiscal position
Romania offers structural cost advantages for deep-tech R&D, combined with full EU and NATO membership for programme eligibility.
Art. 20 CF R&D Deduction
50% supra-deduction on qualified R&D expenses under Romanian fiscal code. Reduces effective tax burden on innovation investment.
REXCD Certification
National R&D performer certification pathway. Enhances eligibility for EU programmes and validates technical capability to institutional partners.
Engineering Talent Pool
Access to top-tier engineering graduates from UPT, UBB, and CUNBM. Competitive compensation vs. Western Europe with equivalent technical depth.
EU + NATO Dual Membership
Full eligibility for DIANA, EDF, EDIP, Horizon Europe, and all NATO innovation programmes. No access restrictions for allied defence work.
Founder
Olos Liviu Virgil
Founder & CEO, Fraktiq S.R.L.Diplomat Engineer from Polytechnic University of Timisoara, graduating with a 9.11/10 GPA in Software Engineering. Over a decade of experience architecting complex systems across full-stack engineering, AI/ML, and distributed systems. Creator of the Fractal Diamond Refinement Process (FDRP) methodology, battle-tested across 64+ production runs and 57 delivered projects. Academic collaborations with Prof. Stefan Oniga (CUNBM, FPGA) and Prof. Octavian Proștean (UPT emeritus, systems engineering).
Ready to build cognitive systems?
Whether you are exploring defence innovation programmes, need cognitive monitoring for critical infrastructure, or want to discuss consortium opportunities — we are ready to talk.