Fast Pipelines Can’t Fix a Flawed Strategy & AI in Software Design

This week Oleg shares a playbook to using leading indicators for successful software, while with Steve we will discuss how AI has impacted software design, and how that may change going forward.

A Ferrari with No GPS: Why Fast Pipelines Can’t Fix a Flawed Strategy
Our industry has fallen in love with the delivery pipeline. We’ve made it faster, slicker, and more automated than ever. The only problem? A faster pipeline just helps you deliver a bad idea to your customers more efficiently.

This talk argues that the real bottleneck isn’t in our tooling, but in our thinking. It’s the classic ‘solution in search of a problem’, the inevitable result of an unseen context gap where critical thinking falters—allowing ‘unknown unknowns’ to quietly blossom into downstream chaos, creating huge cognitive load and causing team thrash.

This session is not a talk about theory or ideal states. It’s a practical, socio-technical playbook for leaders, principal engineers, and anyone trying to drive change from the ‘messy middle’. This playbook isn’t about adding more processes; it’s a set of diagnostic questions and conversational patterns for leaders to use. We’ll explore how to spot the crucial, early signals of success—the qualitative indicators that show you’re on the right track long before the spreadsheets and dashboards turn green.

We’ll dig into real examples of these signals, like subtle shifts in team language or a noticeable drop in ‘developer thrash’. We’ll then connect these human-centric clues back to the hard data we all get measured on, like the DORA metrics. You will leave with an actionable playbook to build momentum and prove the value of your work, especially when you don’t have perfect data yet. It’s about making progress on the messy journey to becoming a data-informed organisation, not just a data-decorated one.

About Oleg (Head of Engineering at Policy Expert)
Oleg Desantis is the founder of Carpe Tempus and a seasoned leader with over 15 years of experience in the software industry, including serving as Head of Engineering - Quality at Policy Expert. He is passionate about creating sociotechnical systems where talented people can actually do their best work, rather than just attend meetings about it. He specialises in leading transformations from the “messy middle” — that chaotic, fascinating space where most of the real work gets done — guiding organisations as they evolve from siloed feature factories into genuinely aligned, outcome-driven teams.

A firm believer in situational leadership, Oleg works to replace organisational ambiguity with clarity and ownership. To help achieve this at scale, he co-created the ‘Initiative Playbook’ at Cazoo — a practical framework designed to answer the age-old question, “Wait, who’s supposed to be doing what again?”.

Outside of tech, Oleg is a passionate scribbler and a father to two young toddlers, who are currently teaching him advanced negotiation and crisis management, mostly through a daily ’trial by fire’. This real-world experience fuels his dedication to helping future leaders navigate the complex, often absurd, challenges of modern software delivery. Ultimately, his work challenges the industry’s obsession with tooling to focus on what truly matters: building the right thing, in the right way. Because a faster pipeline is great, but it mainly helps you deliver a bad idea to your customers more efficiently.

Find Oleg at: Website | LinkedIn

AI in Software Design: Going beyond start-ups
In 2025 we’ve seen whizzy, intricate applications created through AI prompts, with those applications covering great distances with relatively little development effort. All too often though these techniques are unreachable in enterprise IT situations, where legacy landscapes, regulations, and even tooling restrictions hinder adoption and application. When the crushing complexity of an enterprise leaves us working hard to find out what software is actually required, how does AI fit into that world?
Today we’ll take stock as we look at how AI has been impacting our enterprises already, and we will ideate routes to bringing the successes seen in start-ups into the enterprise world. Bring your experiences and take part in how AI is affecting software design in 2025.

About Steve (Software Architect at The Ministry of Software Design)
Steve Read is a software architect, technical leader and consultant, working in both the strategic and tactical aspects of software development. Topics range from large scale enterprise modernisation and transformation down to solution and software design. The common thread across all topics is needing to apply software and systems design principles, dancing between functional and technical considerations to create coherent solutions.

Steve’s background is in Java, having joined the coffee cart back with 1.4. As well as seeing Java develop he has also seen Android develop too, and through enterprise software he has worked through the complexities of developing distributed event-driven architectures.
Steve has defined technical strategy for products, and even whole departments, helping organisations cross the chasm from defining intent to creating an impact.

Steve is an active blogger at The Ministry of Software Design and organiser at Domain-Driven Design London, a socio-technical & systems architecture community.

Find Steve at: Blog | LinkedIn | Mastodon | BlueSky

Recording
Photos and video will be taken at the event. We are intending to publish the presentations including the Q&As (technology gods permitting).

Tickets
This is a ticketed event. Please register for your free tickets in good notice. Please read the terms and conditions.