Zetta Cloud (ZA Cloud SRL) was established in 2013 in Cluj-Napoca, Romania’s “Silicon Valley” and an ideal habitat for fostering technology talent and innovation. The company was built around its founders’ dream to Deliver Actionable Insights through Artificial Intelligence and solve the issue of managing unstructured content – efficiently, accurately, securely and at scale. We offer turnkey Text Analytics software solutions for the Public Sector and Regulated Industries, complemented by AI consulting and research services.
- Pioneers in AI algorithms for fake news detection: TrustServista
- End-to-end Text Analytics product suite covering more than 40 languages: IntelliDockers
- The industry’s only on-premise secure “no-code AI” machine learning platform for business users: Factory No-Code AI
- State-of-the-art Research & Development in AI algorithms for Digital Humanities, Open Source Intelligence (OSINT), Computer Assisted Transcription for Medieval Text Images, Intelligent Knowledge Management Systems.
- Multiple “PatriotFest” awards recognizing our AI innovation achievements for the Intelligence and Security domain.
Read our Manifesto.
No code Machine Learning: Making AI Democratic
Being able to quickly and easily build, deploy and run an AI agent without the need to code, by business users that have no in-depth understanding of AI but do understand the business needs; this is No-code AI, the last mile in making AI democratic.
Meet the Zetta Cloud management team.
Emil, George and Sebastian (Sebi) met at different stages of their career. Emil returned to Romania from Munich in 2007 as an accomplished senior solutions consultant and held various roles as Engineering Manager and Development Manager for software companies with offices in Cluj-Napoca, such as Opware – founded by the famous Marc Andreessen and Ben Horowitz. He then moved onto product management at the world’s leading translation technologies company – SDL – where he met George, a technical solutions consultant spearheading the company’s European growth on machine translation. The two has a common passion: artificial intelligence and making it accessible to the business users. When Emil decided to start Zetta Cloud in 2013, George assisted him with the go-to-market and positioning strategy, joining the company as a shareholder a few years later.
Our journey with AI and Text Analytics began in 2003 when we built Știrili (sounds like “newsly” in Romanian!), an innovate news reading app that learns the users’s preferences automatically and delivers personalized news content collected from the web. We learned how to build web crawlers, AI web content extractors and automatic classifiers. It is a B2C solution available as a web and Androi app that still has thousands of active users. You can try it here: https://www.stirili.ro/
In 2017 we received a Google DNI grant for 50,000 EUR to build a prototype for the world’s first AI-powered online disinformation detection tool: TrustServista. We worked with media professionals and journalists around the world to understand how to get timely access to online news, and how to detect disinformation and fakenews through certain patterns. We incorporated all this knowledge into our software platform and started collecting tens of thousands of article per day. We got praise from Google for our prototype (TrustServista: A question of trust) but we soon realized that the huge amounts of data we were collecting and the complexity of analysis required that we build our own multilingual AI Text Analytics solutions stack: This is how IntelliDockers was born.
By 2019 IntelliDockers was a product with market traction and customers. We started with a couple of languages and several AI engines (summarization, classification, sentiment analysis and named entity recognition), and in only couple of years we reached over 100 engines supporting more than 40 languages. There is literally no NLP tasks that IntelliDockers cannot perform. We chose to deliver our software as self-contained Docker containers with a REST API and a web UI to cater to organizations that value security and performance.
Then our customers came up with a very valid business issue: the out-of-the-box IntelliDockers engines that they were using needed to “learn” their data in order to achieve the highest output quality, or they needed to build new classifiers or add new entities to their engines. And all of this in their infrastructure, and without the support of expert data scientists and machine learning engineers. So we built Factory, the only no-code Machine Learning platform in the world that can run fully on premises and does not require any single line of code to train, evaluate, deploy and run AI text analytics engines.