RadioIntel AI
Automated Radio Intelligence Analysis Powered by AI
Institutional memory for SIGINT operators. Process 50 messages in 15-30 minutes with 90-95% accuracy. Fully offline, tested in combat conditions.
The Critical Problem
Radio intelligence operators face a severe institutional memory crisis
Operator Rotation
3 shifts working day-on, two-days-off. Each shift has its own context. Verbal briefings can't transfer the full scope of discovered patterns, code words, and tactical insights.
Slow Processing
Minimum 1 hour to manually analyze 50 messages. Operator fatigue reduces accuracy to 70-80%. Critical intelligence opportunities are missed.
No Institutional Memory
Impossible to manually compare code words with thousands of previous messages. Connections between events across shifts are lost. Each shift starts "from scratch".
Result: Critical intelligence is lost, response time slows down
RadioIntel AI — Institutional Memory
AI-powered system that preserves and accelerates radio intelligence analysis
Automatic Loading
Intercepted communications via Signal. Automatic PDF processing. No manual data entry required.
AI Analysis
Local language model extracts structured data: call signs, coordinates, frequencies, units, targets, routes.
Vector Database
All messages with context. Full history for every shift. Instant search through thousands of previous intercepts.
Analytics
Search, relationship analysis, activity tracking. Visualize patterns and connections automatically.
Automatically Extracted:
Proven Effectiveness
Tested in combat conditions with verified results
15-30 minutes for 50 messages vs several hours manual analysis
Compared to 70-80% for manual processing of large volumes
Complete autonomy, guaranteed data security, field-ready hardware
Security
100% autonomous operation with no external connections
Institutional Memory
Search thousands of messages in seconds, across all shifts
Technology Stack
Built for reliability, security, and complete autonomy
Artificial Intelligence
Ollama + Gemma2:9b
Local language model, full autonomy
Data Storage
ChromaDB + SQLite
Vector DB for contextual search
Interface
Streamlit + Signal-CLI
Web interface and auto-loading
Hardware
AMD Ryzen 7735H, 64GB RAM
Compact mini-PC for field conditions
Infrastructure
Linux + systemd
Resource management, 24/7 monitoring
100% Offline
Data security guaranteed
Our Team
Combining e-commerce expertise with advanced AI development
Dmytro Cheremnov
Project Lead
- →4 years of IT experience
- →Specialization: Shopify development, LLM integration in e-commerce
- →Expertise: Custom AI applications, customer engagement
Maksym Cheremnov
Technical Lead
- →BSc Computer Science, Czech University of Life Sciences (Prague)
- →5 years full-stack development, last year focused on LLM integration
- →Expertise: Scalable AI solutions, cutting-edge tech with business needs
Our team combines deep e-commerce systems expertise with advanced AI development to create production-grade solutions for high-stakes environments.
Development Roadmap
Quarter 1 (3 months)
Deployment at 6-12 positions • Metrics collection • Model testing
Quarter 2 (3-6 months)
Advanced contextual analysis (RAG) • Predictive analytics • Pattern detection
Quarter 3 (6-9 months)
Mobile dashboard for commanders and intelligence officers
Quarter 4 (9-12 months)
System stabilization & optimization • Extended field deployments • Documentation & training packages • Pre-commercialization preparation
Long-term Perspective
Standardization for Armed Forces • Preparation for commercialization • Export to NATO countries
Ready to Scale
The system is operational. We're seeking accelerator support and strategic partnerships.
Every minute matters. Every detail saved can save lives.
Seeking $90,000 in funding to scale to 6-12 positions and build advanced capabilities. Fund allocation: $50K teams/salary, $37K operations/infrastructure, $3K contingency.
Dmytro Cheremnov — Project Lead