14–15 Nov 2026
Bengaluru
Asia/Kolkata timezone

Unlock the true potential of AI coding agents on Ubuntu, with Workshop

Not scheduled
20m
Bengaluru

Bengaluru

B.M.S. College of Engineering Bull Temple Rd, Basavanagudi, Bengaluru, Karnataka 560004
Talk (30 mins) Desktop

Speaker

Ms Soundarya Rangarajan

Description

For a long time now, we've spoken about model capabilities for the best AI-driven, agentic software development experience: Which model writes better code? Which coding assistant has the best benchmarks?

But in practice, your team likely faces a different (two-fold) challenge: environmental complexity and security.

First, modern software environments are increasingly complex, inconsistent, and difficult to reproduce (remember setting up your dev environment when you first joined the team?). Dependency conflicts, platform-specific tooling and hidden local state often cause more frustration than the coding task itself. Human developers learn to navigate these issues over time. For autonomous agents, though, this means precious space wasted in the context window.

A second, perhaps even bigger challenge, is balancing autonomy of agents vs confidentiality. Where should coding agents run, and what level of access should they have? Yes, I want them to run independently, without asking me for permission to access a directory outside my project in case we need deeper debugging, but I also don't want it reading my ssh keys or bank statements in another folder.

This is exactly the problem space that Ubuntu Workshop is designed to address.

This session is a practical introduction to the newly announced Ubuntu Workshop, and the ideas behind it. We'll look at why giving AI agents their own isolated development environments makes them both more reliable and safer to work with, how Workshop puts a different spin than the sandboxing tools we know already, before diving into a live demo.

You'll see an AI coding agent working inside a Workshop—making code changes, running tests, and debugging issues without touching the host machine. We'll also take a step back to understand the Linux technologies that make this possible, and why they're becoming increasingly relevant as AI becomes part of every developer's workflow.

Prerequisites

None.

Speaker Bio

A classic full-stack dabbler, I’ve worked across domains and technologies, from mobile and game frontends to engineering microservices for travel workflows. Tech wasn’t always the plan though; I once saw coding as an exclusive club I didn't belong to. Community changed that, and today I’m driven by a love for building and enabling others. I write, run workshops, and host hackathons. Placing open-source at the heart of this makes it even better. At Canonical, I help bring open-source tools to builders worldwide.

Summary

The future of AI coding isn't just about better models; it's about better environments. Discover Ubuntu Workshop, see it in action through a live demo, and learn how isolated development sandboxes let AI agents work autonomously without compromising your machine or your data.

Speaker LinkedIn / Socials https://www.linkedin.com/in/soundarya-rangarajan/
Difficulty Beginner

Presentation materials

There are no materials yet.