CTO · AI RESEARCHER · INVENTOR
Strategy that creates patentable inventions. Execution that turns them into valuation.
Strategy first: where a product can be genuinely unique, which inventions to patent, and how that intellectual property raises a company’s fundability and valuation. Then execution: production systems, patent filings, and research published at top venues — ACL, EMNLP, EAMT — with industrial application. That is how I work as CTO, and how I have worked for two decades: patents issued in multiple countries, 25 peer-reviewed papers including journal articles in Constraints and AI Communications, two SaaS platforms licensed to global brands. This site collects that work.
Warsaw, Poland (EU)CTO & Head of R&DInventor · 25 papers
WHY NOW
AI coding agents made software cheap to build. What is left to defend is the invention.
Building is no longer the moatAI coding agents have cut the effort of building software solutions. A feature that took a quarter now takes weeks — for you and for everyone else. Execution alone no longer protects a position.
Patents preserve the advantageA patent on the method or the system is what still holds: it protects the competitive advantage, builds the company’s moat, and raises fundability and valuation. That is where I help — creating the IP and preserving it through patent applications.
Papers build the reputationResearch papers, written from real systems and presented at research and business conferences, elevate brand awareness and reputation — and, ultimately, company valuation.
I am not a patent attorney: I am an inventor of methods and systems in computer science, engineering and artificial intelligence — spanning blockchain and mathematical optimization — and I work with patent counsel on the filings.
Read the note: “The moat moved.” →
WHAT I WORK ON
More on each area →
Strategy first, then execution — what I work on.
AI & innovation strategy
Where a product can be unique, which inventions to patent, and how that IP raises fundability and valuation — plus the roadmap, organization and investor narrative to deliver it.
Invention & IP creation
When AI agents make software cheap to build, the invention is what still protects the moat. Spotting what is patentable, designing inventions so the claims hold, working with counsel through filing in multiple countries — and building the product that embodies them. Three inventions, six filings, patents issued in multiple countries.
R&D & industrial research
Research with industrial application: novel methods, rigorous evaluation, papers at ACL, EMNLP and EAMT — and the patents and products that follow. R&D roadmaps connected to research and technology grants, funding and partnership shaping with US and EU universities.
Research papers & conference talks
Papers written from real systems and presented at research and business conferences — ACL, EMNLP, EAMT, industry summits — elevate brand awareness, reputation and, ultimately, valuation.
AI agents & workflow automation
Multi-agent architectures, RAG and evaluation — benchmarked to beat the baseline and deployed in production — and the automation around them: map the process, automate the repeatable, measure the gain.
SaaS products, browser extensions & desktop tools
From model to a real B2B product, or a focused tool that removes friction: architecture, security, integrations, go-to-market — instrumented so you can see the gain. Two SaaS platforms built from scratch and licensed to global brands.
PROVEN ACROSS EVERY KIND OF ORGANIZATION
I have taken AI from idea to patent or research paper to product — for software houses, consulting firms, startups, my own ventures, and global companies.
SOFTWARE HOUSESETTTechnical project leader — built interactive museum systems with my team; papers at IEEE Digital Heritage 2013 and Museums and the Web 2014.IEEE paper →Talk →MW2014 →ETT →
CONSULTING FIRMSFinsaHead of AI — Data & AI programs for TIM, UniCredit, Vodafone, WindTre.
STARTUPSApp2CheckActing like a co-founder as CTO & Head of Product — two SaaS platforms, an NLP patent, Precisely worldwide partner.
MY OWN INVENTIONSPerfectFinder.ai · TrueFile.aiIdea → patents issued in multiple countries; a patent-pending content-authenticity system.
GLOBAL COMPANIESSumma Linguae → DATAmundiCTO — the award-recognized AIDA Agents platform; AI safety auditing research and system. Clients are primarily big-tech companies, across several of their research and product departments, including frontier labs.
Research published at ACL; products in production for enterprise clients and a global contributor network across 100+ countries.
10/11WMT24++ language pairs where AIDA Agents beat every system
70–98%publication-ready output on industrial data, no human post-editing
14frontier LLMs red-teamed with a reproducible formal-logic method
2B2B SaaS platforms built from scratch and licensed to global brands
SELECTED WORK
All case studies →
Built, benchmarked, patented, shipped.
AGENTIC AI · PRODUCTION
AIDA Agents — a multi-agent LLM platform that beats state of the art
Translator → arbiter/rater → post-editor → re-rater, with RAG. Best on 10/11 WMT24++ pairs; 99.4% terminology accuracy; a Globee Innovation Award for DATAmundi; 2026 AI Excellence Awards finalist; ACL 2026 and EAMT 2026.
Read the case study →
AI SAFETY · RESEARCH → SYSTEM
Formal-logic red-teaming and the AI safety-auditing system
A deterministic, reproducible way to find where LLM safety alignment silently breaks — 14 frontier models, 37.2% bypass — now the core of an AI safety-auditing system.
Read the case study →
INVENTION · ISSUED PATENT
Provably optimal recommendation under hard constraints
A DAG-based optimization engine that returns the mathematically optimal, explainable results under preferences and mandatory constraints — US Patent 12,717,799 B2 (issued 2026).
See the patents →
Emanuele Di Rosa made an outstanding contribution to Stellantis’ benchmarking studies, particularly in data analytics and machine learning. His innovative approach, which includes the application of generative AI to interpret KPIs without hallucinations, provided valuable insights, elevating the quality and precision of the analysis.
Emanuele has a strong background in Artificial Intelligence and applies his knowledge of software engineering and machine learning to address meaningful business challenges. He generates creative solutions to difficult problems, and always accounts for client needs and the customer experience.
More recommendations: a recommendation shared on LinkedIn → · all recommendations on my profile →
IN THE PRESS
All press & media →
