The year 2023 will undoubtedly be remembered as one of tremendous growth in artificial intelligence natural language (NL) technologies. Generative AI and large language models (LLMs) have captured headlines, corporate strategies and legislative agendas around the world, driving development and innovation at an unprecedented pace. As we approach 2024, we want to look back on the year to highlight some of the strides our Team of experts has made by delivering on our mission of “creating technology that transforms language into knowledge and insight.”
The Value of a Hybrid Approach for Natural Language
At a time when the hype surrounding ChatGPT (based on OpenAI’s GPT LLM) has prompted 45% of executives to increase their investments in AI, it’s important to emphasize how organizations can adopt LLMs in a safe, meaningful way. In addition to introducing new features, such as expanded on-premises deployment options and enhanced taxonomy management via third-party knowledge sources and standardized libraries, we integrated OpenAI’s GPT in our platform.
The concept behind our tech enhancements is simple: a single AI approach rarely solves the complex language problems enterprises face. At expert.ai, we uniquely offer the power of the full range of practical AI approaches by combining deep domain expertise with the finest attributes of LLMs, machine learning and symbolic AI—in other words, hybrid AI.
Why Hybrid AI?
According to industry experts like Forrester, “Hybrid AI delivers the best results for natural language processing (NLP) Applications,” and it’s already being used to deliver enterprise value. At expert.ai, through hybrid AI we’re making the potential of natural language data more accessible at the enterprise level, with three goals in mind:
- Continue to design solutions for real-world enterprise requirements and use cases
- Improve our AI core capabilities to deliver the best accuracy, governance and measurable ROI
- Ensure businesses can continue to select from the optimal combination of tools to fit their use case as AI language technology evolves
A Human-Centered Approach is Critical
Earlier this year, we conducted a survey of 300+ business and AI/NL experts from around the world to see how LLMs were impacting the enterprise. Our research revealed that while more than one-third (37.1%) were already considering building enterprise-specific language models, the majority (78%) realized that the efforts required to effectively train a usable and accurate model was a major undertaking.
This is a major challenge that organizations are still contending with. The public domain data used to train LLMs like ChatGPT is not enterprise-grade or domain-specific. Even if a language model has been trained for different domains, it is not representative of what is used in most complex enterprise use cases. In fact, natural language use cases always require some degree of domain-specific training.
The good news is that enterprise models offer many benefits: they can be smaller, more efficient, faster and less resource-hungry while still maintaining high performance. Moreover, a human-centered approach—where subject-matter experts monitor and refine data and inputs throughout the process—will ensure accuracy, transparency and accountability. For more takeaways from our survey, download the ebook “Large Language Models: Opportunity, Risk and Paths Forward”.
Partner to Deliver Value Across Any Industry Process or Business Application
Throughout 2023, we partnered with some of the leading companies in RPA/intelligent automation to help organizations tap into the power of NL capabilities, especially for the Insurance domain. Our new partners include:
Reveal Group. Expert.ai’s unique experience in AI-powered NL solutions complement the Reveal Group’s expertise in intelligent automation services. Our combined strengths support organizations in making RPA bots more effective to immediately accelerate routine tasks and activities, and by enabling the intelligent automation of more complex and strategic language-intensive processes. This includes common cross-industry use cases (email triage in customer services, data analysis, comparison and extraction in legal departments) to more industry-oriented processes (claims management in insurance companies, loan origination and customer onboarding in banking and financial services.)
SS&C Blue Prism. With our combined technologies, organizations can expand their intelligent automation capabilities, delivering new solutions to support strategic business goals. Through data that is accurately analyzed and processed by expert.ai and then automated on the SS&C Blue Prism platform, bots and low-code apps can be extended to more complex processes with high accuracy and lower implementation costs.
Duck Creek. Integrated into Duck Creek’s platform, our solutions help insurers read, understand and extract essential data from medical and accident reports and submissions, score risk engineering reports and triage claims based on urgency and severity, classify records by type, and generate summaries.
The Role of AI in Life Sciences
One of the biggest conversations of 2023 has been the role of AI and NL in the future of Life Sciences. As the amount of pharmaceutical and medical data continues to grow, finding the key insights for the right purpose is critical when it comes to accelerating decision making, improving R&D and bringing medical treatments to market faster.
Our new AI Platform for Life Sciences, launched this spring, helps life sciences companies transform their business and drive more efficiency while preserving data security, privacy and factual accuracy. These enhanced capabilities are honed from real-world implementations, thanks to the many customers who are building their NL strategies and solutions with us, including Sanofi, The New England Journal of Medicine, EBSCO, 3M Health and INSERM.
Our Platform for Life Sciences was selected as a “Best of Show” finalist at the 2023 Bio-IT Awards, the world’s premier event showcasing the technologies that solve problems, accelerate science, and drive the future of precision medicine.
AI for Digital Healthcare
Digital transformation has dramatically increased the volumes of unstructured language data circulating in the healthcare sector. Electronic health records, medical reports, hospital discharge forms, databases, information systems, complaint forms, scientific papers, etc. – represent an exponentially increasing amount of information that can generate valuable knowledge if effectively aggregated and analyzed.
In this scenario, more intelligent solutions can make a crucial contribution. By enabling natural language understanding and processing at scale, AI supports digital transformation with immediate benefits for healthcare professionals, patients/citizens as well as administrative teams at hospitals and healthcare organizations.
In 2023, we focused on the most practical uses cases to accelerate digital transformation across healthcare and patient engagement processes. In Italy, we showcased our enhanced offering at events, such as “AIXA, Artificial Intelligence Expo of Applications” and “FORUM Sanità,” that offered us the chance to meet health domain experts, managers and innovators to learn more about data-driven health solutions and to showcase the benefits that AI solutions offer.
Setting a New Standard for the Insurance Industry
A huge milestone for us this year was bringing the power of generative AI and LLMs to the world of Insurance with the launch of ELMI, the Enterprise Language Model for Insurance. The debut of ELMI followed the release of our AI Platform for Insurance as a further testimony to the unmatched skills and creativity of our R&D, engineering and product teams and to our strong commitment to the Insurance industry behind our go-to-market strategy.
ELMI is a trained language model within the expert.ai Platform for Insurance that provides the generative AI tools insurers need to automate language-driven processes (risk engineering, claims automation, policy review, submissions intake) while offering substantial cost savings.
We were thrilled to see industry recognition of our hard work with awards from PropertyCasualty360’s Insurance Luminaries (Claims Innovation category) and the FinTech Global InsurTech100 list, celebrating the groundbreaking tech companies developing innovative solutions for the insurance industry’s most daunting challenges.
A New Podcast for the Insurance World
This fall, we helped launch the first season of the Insurance Unplugged podcast, hosted by insurance industry veteran Lisa Wardlaw. The podcast features interviews with industry insiders, including Bryan Falchuk, Louis DiModugno, Yannick Even and Ema Roloff who share their insights and best practices on this weekly series.
Subscribe and listen to Insurance Unplugged: Spotify
Financial Services: Enabling True Scalability Across the Most Complex Processes
The ability to identify adverse news, such as investigations, crimes and legal proceedings, from any media and information streams is crucial for making customer and counterparty audits more comprehensive and timely. In today’s increasingly interconnected and global world, financial crimes pose a growing risk on multiple fronts, from protecting the economy, businesses and citizens to safeguarding the integrity of the financial system, social responsibility, corporate reputation and the risk of fines and penalties.
Artificial Intelligence is a valuable ally in supporting financial analysts to manage and make sense of the huge volumes of data available – an activity that is impossible to do manually since these activities require screening thousands of sources every day. In this context, we joined forces with KPMG and the team of process reengineering and automation experts at BCC Banca Iccrea (parent company of the BCC Iccrea Group) to implement an AI-powered news analytics solution to supportg anti money laundering processes.
This solution provides for dual use: it processes specific requests from BCC Iccrea Group banks within anti money laundering (AML) processes (e.g., Know Your Customer and Adequate Verification) and it proactively monitors the entire customer base with alert mechanisms in the event that a target is identified. In this way, AI and human intelligence complement each other and enable true scalability of processes.
Market Recognition and R&D Acknowledgement
As further testimony to our more than three decades of work in the field of natural language AI, we were awarded The Best in Language Intelligence at the prestigious LT-Innovate Awards in September.
We were also included in the annual list of the 100 companies that matter in knowledge management, and our platform was recognized as a Trend-Setting Product of the year. These awards are a testament to the incredible passion and brilliant work that our team of experts perform every day to deliver value and create a tangible impact for our customers.
We would also like to acknowledge the work of our NL scientists, developers and knowledge engineers. Throughout the year, they have participated in several research projects and presented at a number of prominent R&D events across the globe, including:
The 2023 Conference on Empirical Methods in Natural Language Processing. Our research team presented two papers in Singapore under this year’s theme of “Large Language Models and the Future of NLP:” “Multi-word Tokenization for Sequence Compression” and “BUSTER: a BUSiness Transaction Entity Recognition.”
The European Data Conference on Reference Data and Semantics (ENDORSE). During the event, which was organized by the Publications Office of the European Union and Interoperable Europe supported by the Digital Europe Programme, expert.ai participated in the session “Deep Semantic Hybrid Technologies for Information Superiority.” This session focused on how to discover deep semantic attributes from unstructured content and use them within Decision Support Systems (DSS) to support decision making and reasoning.
NOTIONES (iNteracting netwOrk of inTelligence and securIty practitiOners with iNdustry and acadEmia actorS) Events. As part of this EU consortium for furthering research and development in the intelligence community, expert.ai designed and developed a text analysis module for extracting the address IDs of certain types of crypto wallets.
SILVANUS. Expert.ai is among the partners of this European Green Deal Project launched for wildfire management, forest resilience and climate change. Leveraging an integrated technological and information platform for wildfire management, SILVANUS scientists and research engineers have been supporting the civil protection authorities to efficiently monitor forest resources, generate more accurate fire risk indicators, and promote safety regulations.
Looking forward to 2024
As we enter the new year, we will continue innovating to solve real-world problems with the power of AI, accelerating our plan for value creation and empowering enterprises with the most advanced NL capabilities and tools.
Stay in touch with us in 2024! Subscribe to our newsletter and follow us on LinkedIn.
