GIDS news media, articles, insights and virtual events educate and illuminate its audiences so they can be fully prepared to deal with the new realities at work and in their professions.
Get the latest event updates, and insights from today's leading voices.
< session />
Wed, April 28AI-Native SoftwareReliability, Observability & SecurityArchitecture & Distributed Systems
Trust is a fundamental challenge when building AI systems, both ahead of time and while they are running. In this live-coding session, you will explore both sides of the problem by building a safe agent together.
You will examine how a good evaluation pipeline can establish safety ahead of time, and how strong harnesses can provide trust at runtime. Along the way, you will see how these approaches can help create safe AI experiences while still enabling high productivity.
What You Will Learn
How evaluation pipelines can support ahead-of-time AI safety
How strong harnesses can provide trust while an AI system is running
How these approaches come together when building a safe agent
Who Should Attend
Developers, AI Engineers, Software Architects, and technical practitioners building agentic AI systems.
< speaker_info />
Tejas Kumar is an AI Engineer at IBM, based in Berlin, Germany. He has been building on the web for over 25 years, and before IBM he worked at Spotify, G2i, Vercel, Xata as Director of Developer Relations, and DataStax, which IBM acquired. He is the best selling author of Fluent React, published by O'Reilly (ISBN 9781098138714), a book about how React works internally. He hosts ConTejas Code, a weekly long form podcast with 98 episodes. He has given 80 recorded talks at 64 events, starting with a TEDxYouth talk in Doha in 2012 called "Papercuts Can Kill" and speaking at conferences since 2018. He is also an angel investor and advisor to early stage startups, 3 of the 6 he has backed since acquired, by CoreWeave, the Linux Foundation and Supabase, and he backs early stage founders with money, with advice, or with both.
