Pioneering Generative AI Solutions: aiXplain at ADAPT ADVANCE 2024

aiXplain, a leader in end-to-end AI NLP solution development, showcased its expertise at the ADAPT ADVANCE 2024 in Dublin. The event, held at Trinity College, focused on the practical application of generative AI in various industries, with a particular emphasis on large language models (LLMs). Miqdad Dali, Product Manager at aiXplain, presented insights and case studies on the innovative uses of AI.

Case Study 1: AI-Powered Insurance Agent

This case study centered on developing an AI-powered insurance agent to streamline claims processing. The key challenges included:

• Integrating with the client’s existing infrastructure
• Ensuring accuracy and scalability to handle 600 concurrent claims
• Meeting cost, timeline, security, and compliance requirements

aiXplain proposed an AI agent as an efficient solution. This AI agent guided claimants through the claims process by helping them input necessary information, answering their questions, and providing step-by-step assistance. Additionally, the agent detected potential fraud and provided comprehensive claim summaries. The agent also utilized satellite image analysis to assess property damage.

Case Study 2: Collaborative AI for Moxie Robot

The second case study highlighted aiXplain’s collaboration with Groq and Embodied to enhance Moxie, a conversational AI robot designed for children. The project aimed to improve Moxie’s functionality and safety through robust content moderation.

In collaboration with Embodied and Groq, aiXplain provided Moxie access and deployment of content moderation models developed by Groq to ensure safe and appropriate interactions for Moxie, a conversational AI robot for children. Groq, a company that manufactures Language Processing Units (LPUs) designed for high-speed inference of large language models, provided the infrastructure necessary for Moxie’s smooth and fluent conversations.

Solution

• aiXplain played a crucial role in model serving and solution architecture to ensure safe and appropriate interactions for children using Moxie.
• aiXplain significantly improved Moxie’s conversational capabilities and safety.
• aiXplain enabled more natural and impactful interactions for Moxie, primarily through our specialization in hosting and optimizing Large Language Models (LLMs) for Embodied’s young user base. In particular, Groq’s LPU, designed for high-speed language processing, ensured efficient and effective model hosting at aiXplain.

Read the case study

Insights from ADAPT ADVANCE 2024

Anya Belz, Professor of Computer Science at DCU, and Miqdad Dali highlighted the importance of responsible AI development, data governance, iterative project management, and collaboration. They emphasized the growing adoption of AI agents, which hold immense potential for automating complex tasks and delivering tailored solutions.

Notable Attendees and Discussions

ADAPT ADVANCE 2024 featured a diverse range of speakers and attendees, including industry experts, government representatives, and academic leaders. Highlights included:

• Keynote Address: Dragos Tudorache MEP discussed Europe’s position in the global AI landscape and the EU AI Act, emphasizing the opportunity for Europe to lead AI development for the benefit of mankind.
• Director’s Insights: Professor John Kelleher, Director of ADAPT, shared his extensive knowledge on AI adoption, pre-trained models, RAG, prompt engineering, and sustainability considerations.
• Panel Discussion: Chaired by Elaine Burke, featuring Emma Redmond from OpenAI, Medb Corcoran of Accenture, Prof. Dave Lewis of ADAPT, and Dragos Tudorache, discussing the balance between innovation and regulation in AI.
• Innovative Projects: Presentations by Dr. Filip Klubicka, Dr. Lynn Kilgallon, and Dr. Donal Sexton on generative AI’s impact on industry productivity, cultural insights, and medical advancements.
• Start-Up Showcases: Featuring eight start-ups formed in ADAPT research, including eAltra, TiLT, KatoAI, DigiAcademy, Contenseo, VoiceTune AI, aiMapit, and Swan.
• Research Presentations: ADAPT researchers shared their latest projects, with presentations by Dr. Vasu Nedumpozhimana, Dr. Filip Klubicka, Delaram Golpayegani, Dr. Eamonn Kenny, Awais Akbar, and Dr. Martha Lasek-Markey.
• Opening Remarks: The day started with a video from Minister Patrick O’Donovan and a speech by Ciarán Seoighe of SFI, setting the tone for the event.

The event included discussions on the latest advancements in AI, the role of government in AI governance, and the operationalization of generative AI. The event provided a platform for exchanging ideas and exploring collaborative opportunities to advance AI technologies.

Conclusion

aiXplain’s participation in ADAPT ADVANCE 2024 demonstrated their expertise in supporting companies on their generative AI journeys. By presenting practical solutions and addressing real-world challenges, aiXplain illustrated the critical role of data governance and collaborative partnerships in successful AI implementation. Their contributions provided valuable insights into the future of AI in various industries.