About This Webinar 

Many organisations are experimenting with AI, very few are successfully running it in production.

From unpredictable outputs to poor input quality and lack of control, AI systems behave very differently from traditional software, and most teams are not prepared for it.

In this session, we will share practical insights from real-world implementations and explain why AI projects often get stuck between proof-of-concept and production.  

  What you will Learn  

Why AI systems become less predictable in production  

The hidden challenges of real-world inputs  

Why traditional engineering approaches don’t work for AI systems

The role of context, orchestration, and human-in-the-loop design

What teams that successfully ship AI do differently

  Who Should Attend   

  CTOs  

  VP / Head of Engineering

  Engineering Managers

  Product & Technology Leaders

  

Speaker

  

  

Ashwin Bilgi

Engineering Manager

GitLab  

  

Speaker Profile

Ashwin is a Software Engineer and Engineering Manager with 20 years of experience building production infrastructure at scale. He currently builds usage billing infrastructure for GitLab's AI offering, the Duo Agent Platform. Before GitLab, he owned Booking.com's private cloud infrastructure powering their critical production and developer VMs. At Microsoft, he was Principal Engineering Manager for Azure Databricks and Senior SRE at Outlook Mobile serving 130 million monthly active users. Before that, he built a Database platform at VMware. In his free time, he is hacking on VlinderCLI, an open-source tool that helps small teams build AI agents that are easy to debug and repair.  

Built with