Meetup Tomorrow
November
Wednesday
15:00–17:00 2026 · Singapore time
You sign up with the host. foundercal only links to it.
What goes into building better recommender systems, and keeping AI agents from running wild?
Join practitioners from GoTo Group and Docker as they share two different technical perspectives, from using embeddings and knowledge graphs to improve product understanding for search and recommendations, to running coding agents safely in sandboxed environments.
More About The Sharings
Le Yi (Data Science Manager, GoTo Group) & Trieu Le (Algorithm Engineer, GoTo Group) will share on “Building Product Understanding for Search & Recommendations.”
Can embeddings truly understand what a product is? In large-scale marketplaces, noisy and inconsistent product data makes it challenging to build reliable search and recommendation systems.Drawing on GoTo's work, Trieu and Le Yi will explore how LLMs, multimodal encoders, and heterogeneous knowledge graphs can help structure product information, connect product attributes with user behaviour, and build a more reliable semantic layer for search and recommendations. Hear about the approaches and engineering considerations behind making product information more interpretable and useful for downstream ML applications. (Technical Level: 200)
Oleg Shelajev (Developer Relations, Docker) will share on “You're Absolutely Right, It Was Your Home Directory!”
Letting AI agents loose on development tasks sounds great, until one decides to optimise your home directory or break your environment in the process. How do you give agents the freedom to work autonomously without giving them unrestricted access to your system?Oleg will introduce Docker Sandboxes, an isolated execution environment for coding agents, and explore why basic isolation isn't always enough. Through practical examples, he'll unpack how filesystem restrictions, network controls, and secret management can help contain agent behaviour, alongside the engineering considerations behind running agents more safely. (Technical Level: 100)
More About The Speakers
Le Yi is a Data Science Manager at GoTo Group, where she leads the Recommendations Data Science team, focusing on algorithm-driven discovery and personalisation. She brings over eight years of experience developing and scaling data science capabilities across recommendation, pricing, and allocation systems.
Trieu Le is an Algorithm Engineer at GoTo Group, specialising in search and recommendation systems. His work focuses on building scalable semantic layers that transform noisy, multimodal data into structured knowledge for search, recommendations, and downstream ML applications. Previously, he developed AI capabilities for e-commerce at Lazada (Alibaba Group), several of which remain in production today.
Oleg Shelajev works in Developer Relations at Docker, where he focuses on AI-native developer tooling and infrastructure, including Docker Sandboxes, Model Runner, MCP Toolkit, and cagent. A backend developer by training, he has extensive experience in developer advocacy, open-source communities, and software engineering. Previously, he was Head of Developer Relations at AtomicJar, the company behind Testcontainers, which was acquired by Docker, and Lead Developer Advocate for GraalVM at Oracle. His work centres on making complex developer technologies more accessible and practical for engineers.
More About The Series
AI Wednesdays is Lorong AI’s weekly gathering, bringing together practitioners, researchers and innovators for technical discussions on research insights, product development and engineering practices.
Get involved: Learn more about Lorong AI | Speaker Sign-up | WhatsApp Community | LinkedIn | X
Full text on Luma .
Investors in the room
Investors in the room
Founders pitch
