Adam Davis + Juliana Smith

AI makes it incredibly easy to build things. Getting those things to work reliably in the real world is a very different problem.

For September's Practically AI, we're looking at that challenge from two very different perspectives.

Juliana will share her experience building an application with Copilot and Lovable as a domain expert rather than a traditional developer, including what happens beyond generating the code when testing, privacy, GDPR and SEO enter the picture.

Then Adam Davis will take us into the physical world, exploring what happens when AI-assisted development encounters computer vision, physical devices, networks, unpredictable environments and, perhaps most unpredictably, actual people.

Two very different applications, but a similar question: when AI makes building the software easier, where does the hard work move to?

***

## Juliana: Building an Application with Copilot and Lovable: The Easy, the Hard, and the Unexpected

AI can generate applications quickly, but turning generated code into a reliable product requires much more than prompting.
In this session, Juliana shares her experience building PBIX A11y using Copilot and Lovable from the perspective of a domain expert rather than a traditional developer.

We'll explore the easy parts of AI-assisted development, the harder lessons around testing and validating generated code, and the unexpected challenges beyond development, including data privacy, GDPR and SEO considerations.

It's a practical look at what it takes to move from an idea to a working application when AI becomes your development partner.
PBIX A11y is a free tool, and the session is focused on the experience and lessons from building it rather than promoting a product.

***

## Adam Davis: Building Vision Systems with AI in the Real World

AI helps you build faster, but that speed does not guarantee systems that actually work under real conditions.

In this session, Adam will walk through building a real-time, distributed vision system with AI as a co-developer and what happened when it met reality.

Detection shifts under different lighting, timing drifts across devices, and small AI-assisted changes introduce regressions that are not always obvious until the system is already in motion.

This is not a polished success story. It is a look at what happens when something that appears to work is exposed to real-world conditions and starts to behave differently.

We'll run it live using the room, where your phones become part of the system, and you'll see how quickly behaviour drifts once conditions are less than perfect. Adam will then break down what actually made it stable enough to run and the trade-offs involved in getting there.

The focus is not on prompts or models. It is on building systems with AI in the loop, where behaviour is not fully predictable and where you still need to take responsibility for how the system holds up.

AI can accelerate the build, but you still have to make it work.

***

## Who Should Attend

Software engineers, tech leads, architects, engineering leaders, product professionals and domain experts interested in using AI to build real applications.

Whether you're using coding agents every day or exploring how tools like Copilot and Lovable can enable people outside traditional software development to build applications, there should be something here for you.

## Why It's Worth Attending

Practically AI is about the reality of building with AI: real implementations, real problems and lessons you can take back to work.

Juliana and Adam's experiences approach that from two very different directions.

One looks at moving from domain expertise and an idea to a working application with AI as a development partner. The other looks at what happens when AI-generated and AI-assisted software has to survive unpredictable real-world conditions.

Come for the practical lessons and live demo, then stick around for pizza, questions and conversations with other people building with AI across Manchester.

## Agenda

18:30 – Doors, pizza & mingling
18:50 – Introductions & AI Spotlight
19:05 – Building an Application with Copilot and Lovable: The Easy, the Hard, and the Unexpected – Juliana
19:25 – Q&A
19:30 – Building Vision Systems with AI in the Real World – Adam Davis
20:15 – Q&A
20:25 – Open social & refreshments
20:30 – After-event social at The Sawyer's Arms