Data engineering aspects in applying generative AI: The case study of detection of potentially corruptive elements in public procurement procedures

---

Start: 05.02.2025 @ 18:00

---

We invite you to the first PyData meetup of the year. Lecture on Aspects of data engineering in the application of generative artificial intelligence will be held by prof. Eftim Zdraveski, PhD from FINKI, UKIM. The lecture will take place at Base42 on Wednesday 5.2.2025 at 18:00.

Talk:

In this lecture, we will explore the data engineering aspects of applying generative AI to real-world problems, focusing on a compelling case study: detecting potentially corruptive elements in public procurement procedures in Macedonia. Key topics of the lecture include an in-depth discussion of data engineering aspects related to ingesting and processing source data. Likewise, we'll discuss when to utilize traditional databases versus vector stores for storing and retrieving data in generative AI workflows. We will examine the challenges of prompt engineering, mainly when dealing with large document sets and complex question contexts. Special attention will be given to the difficulties of enabling large language models (LLMs) to reason effectively in low-resource languages such as Macedonian, where limited training data introduces additional constraints. To ground these concepts, we will analyze specific examples from public procurement procedures in Macedonia, demonstrating how generative AI can be leveraged to uncover patterns, flag anomalies, and identify potential risks.

Topics:

Explain the data engineering aspects in building applications leveraging generative AI Explain use cases when to use a regular database or a vector store Challenges in prompt engineering related to document and question size Challenges in LLM reasoning in low-resource languages such as Macedonian Examples in the analysis of public procurement procedures in Macedonia

About the speaker

** Eftim Zdravevski Ph.D.** is an Associate Professor and Head of the Institute for Intelligent Systems at the Faculty of Computer Science and Engineering of the Saints Cyril and Methodius University in Skopje, Macedonia. He’s also the Founder of Magix.AI - a company specializing in applied AI and Big Data technologies in various domains. He was recognized as the best Scientist at the Ss. Cyril and Methodius University in Skopje and has published over 170 papers in top-tier journals and conferences in machine learning and Big Data.

He has participated in or led national and international research projects on time series analytics, feature engineering and selection, scalable systems, and ML applications. His active fields of research interest are predictive and prescriptive analytics, FinTech, Big Data, machine learning, data mining, IoT, cloud computing, time series analysis and forecasting, parallel algorithms, etc.

Call for presenters

---

Do you enjoy sharing knowledge and like public speaking? Do you just enjoy sharing knowledge but are unsure how you feel about public speaking? Even if you're not 100% about public speaking, PyData is a very welcoming community and we appreciate any talks about sharing knowledge given by anyone passionate enough to share them.

Sign up to speak on the next PyData:

./speak.sh