Logo Crossweb

Logowanie

Nie masz konta? Zapomniałem hasła

Przypomnij hasło

close Wypełnij formularz.
Na Twój adres e-mail zostanie wysłane link umożliwiający zmianę hasła.
Wyślij

Dive into Data meetup #15

dive-into-data-meetup-15
Wydarzenie:
Dive into Data meetup #15
Typ wydarzenia:
Spotkanie
Kategoria:
IT
Tematyka:
Data:
09.06.2025 (poniedziałek)
Godzina:
17:30
Język:
angielski
Wstęp:
Bezpłatne
Miasto:
Miejsce:
SoftServe Office
Adres:
Q22, al. Jana Pawła II 22, 11th floor
Strona www:
Organizator:
organizator
Agenda:

17:30 – 17:45 Come and grab a drink!

17:45 – 17:50 Official Start

17:50 – 18:20 Talk#1 "Why Delta Lake changes everything?" by Roman Dryndik, Senior Big Data Engineer at SoftServe

18:20 – 18:40 Let’s break the ice

18:40 – 19:10 Talk#2 "A Multi-layered Approach to RAG Hallucination Prevention" by Jakub Mizera, Senior Data Scientist at SoftServe

19:10 - 20:00 Networking and pizza

Opis:

Join us for the 15th edition of Dive into Data – get inspired and explore the latest breakthroughs in data engineering and AI!


When? 09/06/2025, 17:30

Where? SoftServe Office, Warsaw, al. Jana Pawła II 22, 11th floor


Talk #1 "Why Delta Lake changes everything?" Let's learn how Delta Lake transforms chaotic data swaps into high-performing, reliable ecosystems. In this session, we’ll explore how Delta Lake brings ACID (Atomicity, Consistency, Isolation, Durability) transactions, schema enforcement, and time travel to your data architecture—without locking you into a proprietary stack. Whether you’re handling streaming data, prepping for machine learning, or ensuring GDPR compliance, Delta Lake has you covered. 


Talk #2 "A Multi-layered Approach to RAG Hallucination Prevention" This presentation offers practical insights for designing and implementing a robust, factually-grounded Retrieval Augmented Generation system through our multi-layered hallucination prevention methodology. The approach combines five synergistic techniques: advanced query rewriting with Hypothetical Document Embeddings and key term extraction, LLM-powered semantic filtering, automated chain of thought generation, a specialized Guardrails Agent for response oversight, and an agentic dispatcher for iterative answer refinement.


Więcej informacji: https://hubs.ly/Q03mqyYH0

Podobne wydarzenia