Real Projects.
Real Results.
Every project here went from concept to production. No proofs of concept that got shelved. No demos that never shipped. These are live platforms generating real value.
ThreatScoreAI
AI-Powered Vulnerability Risk Scoring
The Problem
Security teams drown in thousands of CVE alerts daily. CVSS scores alone lack context — they don't factor in real-world exploitation likelihood, active threat campaigns, or an organisation's unique attack surface. Teams waste hours manually triaging vulnerabilities that may never be exploited, while genuinely critical threats slip through.
Our Approach
- Built an AI scoring engine that fuses CVSS base scores, EPSS exploitation probabilities, CISA KEV catalogue data, and MITRE ATT&CK technique mappings into a single 0-100 risk score
- Developed a real-time data pipeline ingesting from NVD, EPSS feeds, and CISA advisories with sub-minute latency
- Created a RESTful API with developer-friendly SDKs for seamless integration into existing security workflows
- Implemented intelligent caching and rate limiting to handle enterprise-scale query volumes
“ThreatScoreAI transformed how we prioritise vulnerabilities. We went from drowning in alerts to focusing only on what matters.”
ThreatMonitor.io
Real-Time Cyber Threat Intelligence Platform
The Problem
Organisations struggle to maintain real-time visibility over emerging cyber threats. The National Vulnerability Database publishes hundreds of new CVEs weekly, but most teams discover critical vulnerabilities days after disclosure — often only after they've been exploited in the wild.
Our Approach
- Built a real-time ingestion pipeline from the NVD API that processes new CVE disclosures within minutes of publication
- Developed an intelligent alerting system that notifies teams based on technology stack relevance and severity thresholds
- Created comprehensive search and filtering with full-text CVE description analysis and CVSS scoring breakdowns
- Implemented export capabilities for integration with SIEM platforms and incident response workflows
“We used to check the NVD manually every morning. Now ThreatMonitor does it continuously and only alerts us when something is genuinely critical to our stack.”
Custom Business AI Agent
Private Enterprise Knowledge Assistant
The Problem
A growing professional services firm had critical company knowledge trapped in hundreds of documents, policies, and Confluence pages. New employees took months to become productive, and experienced staff spent hours answering repetitive questions instead of doing billable work.
Our Approach
- Built a RAG (Retrieval-Augmented Generation) pipeline that ingests company documents, policies, SOPs, and internal wikis
- Deployed a fine-tuned LLM on private cloud infrastructure ensuring zero data leakage — no company data touches public AI services
- Created role-based access controls so different departments see only the knowledge relevant to their function
- Implemented conversation memory and follow-up capabilities for complex multi-turn queries
“Our team now gets instant, accurate answers to policy questions that used to require emailing three different people. It's like having a senior colleague available 24/7.”
Ready To Build Your AI Platform?
Talk to SarreniteAI about turning your AI idea into a secure, production-ready business platform.