Generative AI is on everyone’s lips, but it is not that widespread in companies. There’s a lot of talk about it, but moving from words to deeds is a whole different story. Alongside the fascinating promises of this new frontier in the evolution of artificial intelligence, there are doubts and concerns, above all because public solutions are not “enterprise grade”, that is, they do not guarantee the reliability and levels of security that companies cannot (and they don’t want to give up. The path, on which Appian has already declared itself, lies in Private AI. We talked about it with Lorenzo Alegnani, Vice President Customer Success Southern Europe di Appian.
Lorenzo Alegnani, Vice President Customer Success Southern Europe di Appian
T4F: On the subject of Artificial Intelligence, what is the approach that companies are adopting? What are the elements at the center of their discussions and decisions?
Lorenzo Alegnani: I could summarize the answer in a sentence as simple as it is explanatory of the current context: “there is a lot of interest but still a lot of uncertainty about how to really apply technologies based on artificial intelligence, as well as concern about possible impacts”.
Companies are asking for help to understand how they can solve any problems or inefficiencies, regardless of the technological tool, even if solutions based on artificial intelligence arouse a lot of curiosity and interest at the moment. Companies need to find valuable solutions that allow them to have a competitive advantage, greater productivity, better efficiency… this never changes. The difference is that now there is an almost infinite panorama of technological offers and this destabilizes them, they first of all need support to clarify and understand what can really make the difference.
T4F: Today there is a lot of talk about Generative AI but companies are making different and less “trendy” choices compared to what we read in the media. Looking specifically at this latest wave of innovation in the AI field, in your opinion what do companies really need and how could they best exploit generative artificial intelligence?
Lorenzo Alegnani: Even in this specific case, it is important to provide companies with the right support to understand new technologies and enable them to produce value thanks to them.
Speaking specifically about Generative AI, there are companies that are introducing it as a building block for the automation of certain processes, others to accelerate the release of services with the generation of software code, still others to improve discovery and knowledge base systems… there is never a single recipe, much less a common path. It is never the tool that defines the path (much less the objectives).
What we observe directly, for example, is a strong interest in facilitating and accelerating the response to the customer in the customer care area, simplifying and speeding up the discovery and aggregation of information and, lastly, simplifying and releasing new applications more quickly through functionalities of AI to interpret natural language and translate it into code.
It is important to emphasize that companies do not run after consumer technologies. They are certainly influenced in their interest, but there are aspects of effectiveness of the tools, reliability and security performance, privacy, data protection and compliance (regulations but also internal policies).
Just think of the recent questions opened by the Guarantors of some European Union countries which raise some concerns about possible data violations by players offering public AI systems for model training. But this is just one of several concerns raised by governments and companies. Generative AI of a public nature, like the one behind the services we have heard so much about in recent months [ChatGPT di OpenAI, solo per citare il nome più “rumoroso” – ndr]at the moment it is arousing interest but also a great deal of concern in corporate environments (public administration and private companies).
At Appian we have a precise vision. Our strategy to make AI private follows methods of building and deploying AI technologies that respect the privacy and control of user and organizational data. We want to give concrete answers to companies and allow them to truly reap the benefits of AI by overcoming the barriers of public artificial intelligence and preserving the security and intellectual property of our customers.
T4F: In Appian’s strategy, we talk about the democratization of Artificial Intelligence. What does it mean?
Lorenzo Alegnani: When we talk about AI democratization we are referring to the process necessary to make Artificial Intelligence accessible and usable by a wide range of users, including those who may not have advanced technical training. This means, for our part, concentrating efforts towards the development of our platform and the tools we integrate with a particular eye on simplifying use, which does not necessarily mean having banal or simplistic tools. Appian increasingly wants to “simplify technological complexities” by making the tools available to users easy and intuitive.
Appian has always been a software company that offers a business process automation platform that allows organizations to automate and streamline their workflows and business processes. We are certain that AI plays a central role in process automation, and we have integrated it into our platform and made it accessible to companies and users (always in that Private AI logic which takes into account the very important aspects of security and reliability ).
T4F: From a broader point of view, how does AI tie into your Process Automation-based model?
Lorenzo Alegnani: To answer, I’ll give a concrete example, taking up the threads on the subject of Generative AI.
Public AI models train themselves using large data sets. Large language models (LLMs) like ChatGPT train their dataset on a wide range of data from the Internet, with a non-trivial problem in terms of effort, as well as security (you have to sift through a lot of noise to arrive at the correct data).
Appian’s private AI models are trained only on company data. For example, to develop a system that automates the document identification and classification process, you can quickly train a model using real company documents as a representative dataset. The implementation is then accelerated thanks to low-code design and Generative AI to produce software code, features present natively or already integrated into our platform. Appian AI Copilot uses generative AI to accelerate low-code development. The result is an artificial intelligence model that automatically classifies incoming documents. This model remains private and generates much more accurate results and classifications, because it is developed and trained on specific, real company data. This is what companies ask for, this is what they need. And it is not always necessary to implement pharaonic projects. Appian’s methodology also allows us to identify and implement projects with setup and go live in just eight weeks. In our roadmap there is an intense path of continuous development and technological integration to add advanced features (including Generative AI) to our platform, always keeping the pillars of our positioning firmly in place, which does not change. We act on Process Automation. It’s our DNA.
breaking latest news © (Article protected by copyright)