The One About Evals (Round II)
As LLM systems move from demos to production, evaluation becomes essential for understanding what works and what improves over time. Explore how to build evals that guide iteration and inform system design, using lessons from real-world LLM and RAG deployments.
More About the Sharings
Gabriel Chua (Developer Experience APAC, OpenAI) Evals in Practice: Building Feedback Loops for LLM Systems
Gabriel will share on how to design and run evaluations that produce reliable signals for improving LLM systems. Explore concrete ways to get started, common failure modes, and patterns that hold up in production. Learn more about how evals connect to prompt design, system architecture, and fine-tuning decisions, with an emphasis on sustained feedback loops. (Technical Level: 200)
Aritejh Goil (Senior Software Engineer, Visa) RAG Evals & Making the Case for Them
Building production-grade RAG systems and chatbots is only half the battle, proving their effectiveness and reliability is just as critical. Aritejh will share practical insights from deploying GenAI systems at scale, focusing on evaluation frameworks for RAG pipelines and chatbot performance. He will then also cover how to engage stakeholders and make the case for robust evaluation processes, ensuring that the right metrics are in place to align technical teams and leadership. (Technical Level: 200)
...and more to come! keep a lookout! 👀
More About the Speakers
Gabriel Chua is a Developer Experience Engineer for APAC at OpenAI, where he works with developers to build, ship, and scale production AI applications. Previously, he worked as a data scientist on MLOps, LLM solutions, and applied Responsible AI, including building ML and LLM systems to combat online scams. He is active in the AI community, previously co-organizing AI Wednesdays and other practitioner-focused events, and holds degrees from the London School of Economics and MIT.
Aritejh Goil is a Senior Software Engineer at Visa, building production-grade GenAI systems including custom RAG pipelines and chatbots using LangChain and Azure OpenAI. He has optimised deployment pipelines and LLM performance while developing sentiment analysis and recommendation models for enterprise applications and has contributed to IMDA's GenAI evaluations and Red Team efforts. Currently pursuing a Master's in Computer Science at Georgia Institute of Technology, Aritejh holds a Bachelor's in Engineering Science from NUS. His interests include all things GenAI and new ways to improve the harness and infra around generative AI applications.
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