Name: #47 Allegro Tech Talk - AI edition
Start: 2026-03-17T18:00:00+01:00
End: 2026-03-17T20:00:00+01:00
Location: Allegro, Wierzbięcice 1B, Poznań

➡ Rejestracja: https://app.evenea.pl/event/allegro-tech-talk-47/
Zapraszamy Was na #47 wydarzenie z serii Allegro Tech Talk, podczas których dzielimy się wiedzą, wzajemnie inspirujemy oraz integrujemy przy dobrej kawie ☕, napojach🥤 i pizzy 🍕. Ten meetup odbędzie się w Poznaniu, w biurze Allegro i jest częścią serii spotkań, na których rozmawiamy o tematach związanych z AI. Nie zabraknie dobrych historii, praktycznych przykładów i ciekawych insightów. Do zobaczenia 17 marca 2026 roku o 18.00!
Temat wydarzenia - Allegro Tech Talk - AI edition
Data: 17.03.2026 (wtorek)
Godz: 18:00 - 20:30
Offline: Poznań (biuro Allegro - ul. Wierzbięcice 1B, kompleks biurowy Nowy Rynek) - do zapisanych osób prześlemy wszystkie wskazówki logistyczne
AGENDA
18:00 - 18:10 - Rozpoczęcie
18:10 - 19:00 - Marcin Wiktorowski - Pixels, Vectors, or Text? Allegro Search Engine Goes Multimodal
19:00 - 19:10 - Przerwa
19:10 - 20:00 - Riccardo Belluzzo - Beyond the Search Bar: Engineering the Future of Conversational E-commerce
20:00 - 20:30 - Zakończenie & networking
➡️ Pixels, Vectors, or Text? Allegro Search Engine Goes Multimodal
Why does a search for "red floral dress" sometimes return a red dress with no flowers, or a floral dress that isn't red? Because our search engine is currently "blind" - it reads text but ignores the image.In this deep dive, the Computer Vision Team will open the hood of our latest initiative: fusing computer vision with semantic search. We'll skip the hype and look at the hard engineering problems: How do you add images to a two-tower retrieval model without destroying latency? Is it better to train on raw pixels or use VLMs to describe images for us? We will review scientific literature, share our three core approaches, and discuss the complex reality of "fusion" - making text and images agree on what a product actually is.
➡️Beyond the Search Bar: Engineering the Future of Conversational E-commerce
As consumer expectations shift toward personalized, intuitive interactions, traditional keyword and vector searches are becoming insufficient. This talk explores the evolution of e-commerce platforms from standard "retrieve and rank" models to dynamic, conversational interfaces. We will dive into the technical architecture required to build these systems, moving beyond basic chatbots to sophisticated Agentic Workflows. We'll also examine the critical roles of LLM orchestration, context engineering, and system evaluation, illustrating why a modular, multi-agent approach is now the gold standard for achieving scalability and reliability.
W spotkaniu udział wezmą:
Marcin Wiktorowski
Marcin is a Research Engineer in the Computer Vision team at Allegro Machine Learning Research. His research centers on multimodality and unbiased evaluations. He is deeply curious about the world, and when he's not tuning loss functions, he is usually contemplating the philosophical implications of everyday life.
Riccardo Belluzzo
Riccardo is a Senior Research Engineer at Allegro, working in the Language Intelligence division. Over the past few years, he's been involved in a wide range of projects, from training multilingual language models to developing representation learning for e-commerce. He's also recently led several successful projects, bringing LLM-based solutions into production by applying the latest NLP breakthroughs. Riccardo is also a dedicated member of the local ML community, sharing his expertise at meetups and conferences. When he's not researching, he enjoys playing electric guitar, cooking delicious Italian meals, and exploring the world.
Do zobaczenia w Poznaniu!