Clariter — Cognitum Social Analytics
Multilingual sentiment analysis across six languages and the major social platforms — millions of posts processed with 95%+ classification accuracy.
My work
Where it started: social and news analytics, credit quantification — language models put to commercial work.
Multilingual sentiment analysis across six languages and the major social platforms — millions of posts processed with 95%+ classification accuracy.
Media intelligence platform associating companies with news coverage in real time — over 1 million articles processed monthly with 88% association accuracy.
AI system that reads court and financial documents to quantify justice-related credits — 94% accuracy in legal entity extraction and 70% faster assessments.
Specialized data science consulting for enterprise AI delivery — custom algorithm development, integration strategy, and hands-on enablement of client teams.
Registry AI, citizen platforms, tender analytics and national data infrastructure with Sogei and Infocamere.
Automated extraction of corporate data from assembly minutes and registry filings for the Italian Chambers of Commerce — 92% extraction accuracy and 60% faster processing.
Guided, mobile-friendly redesign of Italy’s real estate registration workflow — user completion rates up 65% and support call volumes measurably down.
Digitization platform improving how citizens interact with Italian public services across web, mobile, and kiosks — with accessibility designed in for every demographic.
Analytics platform that mines tender and procurement documents to map competitive positioning and predict bid success probability for smarter bidding strategies.
Large-scale data and analytics platform for the Italian government’s IT agency, consolidating national data sources and powering predictive analytics for public service planning.
The longest thread: from mail dispatching to fraud detection to the Generali document-intelligence flagship.
Prototype that classifies and routes insurance customer emails by policy type and customer segment — 87% classification accuracy and 65% faster routing.
Automated processing of certified email (PEC) and broker documentation for a leading Italian insurer — 100% regulatory compliance, 96% extraction accuracy, and 85% less manual handling.
Intelligent email classification and dispatching platform for an insurance operator, integrated with claims management systems — cutting manual mail handling by 70%.
Deep optimization of a core insurance data platform — smarter relationship modeling, indexing, and archiving strategies that significantly cut query times and maintenance overhead.
AI engine that scores insurance claims and surfaces anomalous patterns for investigators, replacing static rule-based screening with adaptive machine learning.
Multi-year AI platform for one of Europe’s largest insurers, classifying and routing millions of emails and documents across business units. Reached 97% document classification accuracy and cut manual processing time by 80%.
Risk assistants, chemical formulation AI and process optimization on the factory floor.
AI-powered risk assessment for a global industrial automation leader — real-time monitoring across manufacturing operations that cut risk incidents by 40%.
Automated CV parsing and candidate-job matching for an HR services group — 93% extraction accuracy across multiple languages and 80% faster screening.
AI platform that suggests optimal polymer formulations and predicts compound performance before lab testing — cutting formulation development time by up to 50% and reducing material waste.
Machine learning system tuning centrifugal decanter operations in real time — 15% better separation efficiency with less operator intervention and more consistent product quality.
The ENEL saga and Terna's enterprise enablement: capital, trading, service and safety.
Predictive models optimizing cash flow and MRP parameters for Italy’s largest energy company — 15% improvement in cash flow efficiency and 12% lower inventory holding costs.
Robotic process automation across energy trading back office operations — 85% of routine settlement and reporting processes automated, with 70% faster processing and fewer errors.
Year-long AI training and experimentation program for Italy’s national electricity grid operator — 25+ engineers enabled and 5 proof-of-concept applications built for grid management.
Enterprise chatbot handling customer service for ENEL’s advanced energy services division — 85% autonomous query resolution, 24/7 availability, and 40% lower service costs.
Computer vision system analyzing live video feeds across energy infrastructure to flag safety violations as they happen — 92% detection accuracy and 60% faster incident response.
Let's talk