XMANAI partner spotlight – Suite5

Fenareti Lampathaki_Suite5

Name: Dr. Fenareti Lampathaki

Job title: Technical Director

Organization: Suite5 Data Intelligence Solutions Limited

Bio: Fenareti Lampathaki (female) holds a Ph.D. Degree and a Diploma – M.Eng. Degree in Electrical and Computer Engineering (Spec.: Computer Science) from the National Technical University of Athens (NTUA) and M. Sc. Degree in Techno-Economics (M.B.A.). She is a co-founder and Technical Director of Suite5, fascinated with how data intelligence can be leveraged into everyday operations of any stakeholder, enabling them to reach better decisions at the right time. During the last 17 years, she has successfully led the team’s research and management activities in a series of commercial and EU-funded R&D projects in multiple domains related to data interoperability, modelling and analytics/AI. She is currently the Technical Coordinator of the XMANAI Project (H2020 ICT-38-2020) and the SYNERGY Project (H2020 DT-ICT-11-2019, “Big Energy Data Value Creation within SYNergetic enERGY-as-a-service Applications through trusted multi-party data sharing over an AI big data analytics marketplace”).

Name: Evmorfia Biliri

Job title: Head of Data Science

Organization: Suite5 Data Intelligence Solutions Limited

Bio: Ms. Evmorfia Biliri holds a Dipl.-Ing. in Electrical and Computer Engineering from National Technical University of Athens (NTUA).  She has a strong background in information systems and web technologies, with a focus on text mining and data analysis. She has been working in the Data Science team of Suite5 since 2018. Previously she worked in the Decision Support Systems Laboratory EPU-NTUA. Her research interests lie in exploratory data analysis, query intent mining and explainable machine learning. She has experience with many Python frameworks for large scale data processing and machine learning pipelines. Over the last eight years she has been involved in various EC co-founded projects, including AEGIS (Advanced Big Data Value Chain for Public Safety and Personal Security), ICARUS (Aviation-driven Data Value Chain for Diversified Global and Local Operations), SYNERGY (Big Energy Data Value Creation within SYNergetic enERGY-as-a-service Applications through trusted multi party data sharing over an AI big data analytics marketplace) and XMANAI (eXplainable MANufacturing Artificial Intelligence).

Q: What is your organization’s role in XMANAI?

A: Suite5 is the Technical Coordinator of the XMANAI project and leads the design and development of core parts of the XMANAI Platform, namely the Explainable Artificial Intelligence Services (WP3) and the Data Asset Sharing Services. Suite5 is also responsible for the consolidation of the XMANAI MVP (Minimum Viable Product) and the overall platform reference architecture while being heavily involved in the delivery of Explainable AI models to solve concrete manufacturing problems (in WP4). Finally, Suite5 supports the Whirlpool demonstrator to leverage the XMANAI offerings in order to address the underlying problem of “Product Demand Forecasting”.

Q: How does XMANAI relate with your or your team’s background and interests?

A: Suite5 delivers end-to-end data-driven intelligence solutions and architectures to clients from different domains with the aim of transforming their latent and unexplored data assets into actionable insights through AI. Thanks to its focus on the explainability aspects of AI to bring humans in the loop, XMANAI is placed at the core of both the background and the interests of Suite5.

Q: What is the novelty of XMANAI and the main benefits envisaged for your organization?

A: XMANAI aims at bringing AI closer to humans by introducing explainability across the data/AI lifecycle (from the early experimentation to the production phases) and by applying explainability over the input datasets, the ML/DL models and the acquired results/predictions from a pipeline. XMANAI explores how AI can be made more understandable to different end users from the manufacturing domain which is very strong pain point in reality in any industry today.

The benefits and added value that Suite5 expects to acquire through its participation in XMANAI concern the acquisition of further know-how on Explainable AI, as well as bringing to the market and commercializing new advancements, products and services that emerge from XMANAI. Moreover, during the project, Suite5 will explore how explanations can become an integral part of AI and how collaboration among different data scientists, as well as with domains experts, can be achieved in a fruitful way and channel the knowledge and experience gained into its commercial offerings and products.

Q: Which target groups can benefit from XMANAI?

A: XMANAI is designed to decrease the barrier entry for manufacturers in the AI era by solving concrete manufacturing problems accompanied by solid, understandable insights into the why and how any decision of AI models is taken. To this end, XMANAI counts on the collaboration among different actors (data scientists, data engineers and business users with different profiles) who benefit from XMANAI in different ways when configuring, running and applying the results of AI pipelines.

Q: As the technical coordinator, how do you envision the XMANAI progress beyond the state of the art?

A: XMANAI brings together a number of cutting-edge technologies (ranging from AI pipeline design and execution, MLOps, and visual exploration, to data management, security, and sharing), to contribute to its mission to “make AI understandable”. Although such underlying technologies currently face their own challenges and pain points that offer fertile ground for technological progress, XMANAI focuses on Explainable AI as its endgame and expects to go beyond the state-of-the-art at technology and manufacturing innovation level. From a technology perspective, XMANAI shall deliver a platform that allows the “collaborative” design, experimentation, and deployment of Explainable AI pipelines, that contain elements/tasks that are made explainable by design: raw datasets structure and semantics, ML/DL models inner workings, tailored explanations for the pipeline results per target audience (at global/local level). From a manufacturing innovation perspective, XMANAI shall release a Catalogue of Explainable AI Models that combines AI models with their associated explainability techniques and are offered in an out-of-the-box manner to solve different (manufacturing) problems. XMANAI will also provide a set of manufacturing apps that interact with the XMANAI Platform to provide targeted dashboards with insights to its demonstrators.