Damien Ryan - Working Sanely to Deliver Small Bits of Value

Summary

DevOps Shorts – Ep. Damien Ryan: Working Sanely to Deliver Small Bits of Value

Host: Ant Weiss
Guest: Damien Ryan (Director of Engineering, Featurespace)
Date: May 3, 2020
Length: ~15 minutes

---

Overview

This episode of DevOps Shorts features Damien Ryan, whose career in tech spans over two decades, from hands-on applied science to leadership at Featurespace (a company specializing in adaptive behavioral analytics for fraud detection). The conversation explores Damien’s passion for IT, the pivotal moments that shaped his approach to DevOps, and his vision for the future of the industry. True to the show’s format, the discussion is concise but packed with insight, focusing on how working sanely and delivering incremental value leads to better outcomes for teams and customers.

---

Key Discussion Points & Insights

1. Why Damien Loves Information Technology

• Childhood Fascination:
  Damien’s journey began with an AMSTRAD computer, delighted by how simple commands produced exact results—something people often didn’t do.
• Purpose-Driven Work:
  Over time, he realized the power of IT to solve real problems, particularly at Featurespace where the mission is to “find fraud that will stop real people losing money.”
• [02:14] Quote:

  “Started off liking it back when had an AMSTRAD computer and you could type in a few things and did exactly what you wanted, but people would just not do that for you. Where I started to love it was realizing that you could take all of the fun stuff and actually solve real problems for people... we're trying to find fraud that will stop real people losing money.” — Damien Ryan

---

2. Damien’s DevOps “Aha” Moment

• Painful Early Days:
  Software delivery meant painstaking, manual processes—zipping up files, shipping via FTP, manual testing, hoping nothing broke.
• Shift with XP and Automation:
  Exposure to Kent Beck (creator of Extreme Programming) and the XP philosophy was a revelation:

  “If you ask the customers what they want and give it to them, they might actually like the software you give them and stop yelling all the time.” — Damien Ryan, [04:10]

• Beyond Tools to Organizational Change:
  Automating deployments brought relief, but the “aha” was realizing the real challenge was organizational and cultural change, not just technical improvements.

  “To really kind of make change across the business, you kind of have to dig into more than just another framework. And it's how we work is a bit that really excites me more than what we actually do.” — Damien Ryan, [05:32]

• DevOps as Sanity & Reproducibility:

  “Sanity, reproducibility. Just making things work so we don't all have to die doing it.” — Damien Ryan, [06:19]

---

3. The Future of DevOps and IT

• Moving Beyond “DevOps” as a Job Title:
  Damien critiques the trend of “DevOps” becoming siloed as a role—“squashed by a bunch of consultants trying to make a place for themselves.”
  • Advocates for cross-functional teams collaborating on real value delivery.
• Small Batches, Real Value:
  Envisions a future where “people working sanely to produce small bits of value quickly and safely to customers” becomes the norm across IT.

  “The industry [should be] moving around value… having those customers have their real problem solved as soon as possible.” — Damien Ryan, [08:00]

• Not Yet Evenly Distributed:
  Despite progress, the principles aren’t universal:

  “A lot of people doing DevOps and Agile… they're doing it quicker, but they're not actually producing real value because you still have angry customers who don't get the things that they want.” — Damien Ryan, [08:47]

• What Happens After DevOps?
  Predicts that, eventually, “DevOps” will just be software development—delivering small increments of value, based on actual user needs, and working together sanely.

  “DevOps is kind of the state we have to move through to get to the place where software development is delivering small pieces of value based on what real people want to do.” — Damien Ryan, [10:01]

---

Bonus Lightning Questions & Memorable Moments

What is Adaptive Behavioral Analytics?

• Key Difference:
  Models that “retrain based on data that comes in,” adapting quickly to new fraud patterns as behaviors change.

  “It's not just that initial training set that works as data... the models themselves can adapt and reuse those new data to retrain, catch different types of fraud as they come in.” — Damien Ryan, [11:24]

Favorite Programming Language?

• PowerShell:

  “Still love PowerShell. If I ever get to... program again or do anything, I usually use PowerShell mainly because it's what I'm comfortable with and it just makes it really easy.” — Damien Ryan, [12:32]

Most Hated Language?

• Bash:

  “Love, hate relationship, really, but it has to be Bash... Bash gets things done, but it's also extremely ugly and can explode if you look at it the wrong way.” — Damien Ryan, [12:50]

  • Ant: “Yeah, that's the reason to love something also... this feeling of danger.”

Great Management & Team Dynamics

• Best Managers:
  Influential managers never felt “one-sided,” always working “eye-to-eye,” like the Lennon/McCartney creative partnership.

  “The key thing... has always been that it's never felt like a one-sided relationship... They manage by showing you where to go and you don't even know you're being managed at that point.” — Damien Ryan, [13:20]

• Communication as the Secret to Great Teams:

  “Engineering is a team sport, which, you know, engineers aren't great at.” — Damien Ryan, [14:50]

Why Move to the UK?

• Economic Context:
  A career move during an Irish recession led to a short-term UK job… that turned into 18 years and counting.

  “This one came up in the UK. I thought let's give it a shot for a couple of months, see how it goes. And that was 18 years ago.” — Damien Ryan, [14:11]

---

Notable Quotes & Timing

| Timestamp | Speaker | Quote | |-----------|---------|-------| | 02:14 | Damien | “Started off liking it back when had an AMSTRAD computer… you could take all of the fun stuff and actually solve real problems for people… we're trying to find fraud that will stop real people losing money.” | | 04:10 | Damien | “If you ask the customers what they want and give it to them, they might actually like the software you give them and stop yelling...” | | 05:32 | Damien | “To really make change across the business, you kind of have to dig into more than just another framework. And it's how we work is a bit that really excites me...” | | 06:19 | Damien | “Sanity, reproducibility. Just making things work so we don't all have to die doing it.” | | 08:00 | Damien | “The industry [should be] moving around value… having those customers have their real problem solved as soon as possible.” | | 08:47 | Damien | “A lot of people doing DevOps and Agile… they're doing it quicker, but they're not actually producing real value…” | | 10:01 | Damien | “DevOps is kind of the state we have to move through to get to the place where software development is delivering small pieces of value based on what real people want to do.” |

---

Timestamps for Important Segments

• [02:14] — Why Damien loves IT
• [03:46] — Early software delivery pain & the DevOps "aha" journey
• [07:27] — Hopes and predictions for the future of DevOps
• [11:24] — Defining adaptive behavioral analytics
• [12:32] — Favourite and most hated programming languages
• [13:20] — Great management principles
• [14:11] — Career moves and working in the UK
• [14:50] — The secret to great engineering teams

---

Overall Tone and Takeaway

The conversation is energetic, reflective, and deeply practical. Damien advocates for a humane, communication-centered approach to engineering work, with a focus on small batches, meaningful automation, and delivering real value—always striving for sanity over heroics. The episode embodies its own ethos: insightful delivery in a small, digestible batch.

Transcript

[A]

DevOps Shorts. DevOps Shorts. The show to listen to when you DevOps hurts. And even when you're going strong, it's short and sweet. So come along. Hello folks and welcome to DevOps Shorts. The show where we invite wonderful human beings to have a lightning fast conversation about devs, orbs and other mythical creatures. The show where each episode only lasts 15 minutes and we're focused on asking only three questions. So it's short and sweet. Why? Well, because if there's one thing we all know is, it's that great delivery comes in small batches. Right? And our guest for tonight is Damian Ryan. Damian has spent 22 years working in tech and going from applied science through all kinds of build and release engineering and management roles to his current position which is director of engineering at featurespace, a company that builds adaptive behavioral analytics technology for fraud and financial crime management. I got it from your website.

[B]

Okay.

[A]

Damian, how are you today with the lockdown and everything?

[B]

It's day 33. Forgot how to actually be human anymore. Suddenly realizing I'm an extrovert. Out of all the things I thought I was extrovert was dot 100 discovery late in life.

[A]

Cool. We're all finding out new things about ourselves. Okay, let's dive straight into the questions without wasting any more time. So our first question is, why do you love information technology? I have this stupid, probably naive assumption that you love what you do and I'd love to hear why. What happened in your life that made you love it? Maybe you don't love it actually, maybe you hate it. This won't be a love story. This will be a hate story. But it's your story. That's what important. And I'm starting the timer now. Go.

[B]

So, started off liking it back when had an AMSTRAD computer and you could type in a few things and did exactly what you wanted, but people would just not do that for you. Where I started to love it was realizing that you could take all of the fun stuff and actually solve real problems for people. So for instance, like what we do at Featurespace, we're trying to find fraud that will stop real people losing money. Okay.

[A]

And that's it. Okay, cool. That was very short. Much shorter than I thought. Okay, we have leftover time. Great. Just as we said, there are more questions in the backlog if we do it faster. Okay, so our next question will be. What was your DevOps Aha moment. In DevOps Handbook, the book that every one of our listeners should probably read, each of the authors describes their DevOps Aha. Moment. Right? That situation in their lives that made them realize that the way things work in it sucks, but there's a better way. What was that moment for you? Maybe not a moment. What was that situation? How did you come to realize that there is a better way and maybe DevOps it is and your time is starting now.

[B]

So when I started all those years ago, delivering software was basically packing up a bunch of zips, a bunch of files into a zip and sending that over an FTP to a customer. And that would take a week together. All the branches from all of the people who are involved in software, testing it all manually and making sure that would get to the customer.

[A]

Sounds fun.

[B]

This was not fun. It was busy work. It kept me in Smarties. But you know, there had to be a better way. And that's when Kemp Beck started banging his drum. He came along to the company, talked about xp and then something went off my head. I was like, hang on. If you ask the customers what they want and give it to them, they might actually like the software you give them and stop yelling all the time. So gradually kind of move from moving files around to sending installers about. Then started working on Webster and similar process, went into a company now updating the software, was copying a bunch of files around, restarting some services all manually and hoping it worked. And what is he just set? It's crazy busy work. Great for keeping people in jobs, not great for their sanity. So gradually started automating things and realizing that the more you can automate and get out of the way, the more interesting things you can look at. And that's when DevOps came along. And I used to call it build engineering. I think it used to be that for a long time. And then DevOps came along and it just got so complex and so different very quickly. So then after, after a while the problems stopped being technical for me. It's easy to learn new things and new languages and all that, but to really kind of make change across the business, you kind of have to dig into more than just another framework. And it's how we work is a bit that really excites me more than what we actually. And that's where I am right now. Just trying to make sure that we can work sanely and effectively without heroes and without heroics.

[A]

Okay. Preventing fraud and financial crime. Right.

[B]

Without having to work 24 hour days and have crunches. Okay. Okay, good.

[A]

So DevOps for you is sanity, right? That would be a correct definition.

[B]

Okay. Sanity, reproducibility. Just making things Work so we don't all have to die doing it.

[A]

Okay, great, wonderful. And on to the next question. We're doing great here. So that's the most exciting question for me because I tend to introduce myself as a software delivery futurist. I like looking into the future. The future excites me most of all. And now this is your chance to look into the future and tell us what you think is next for DevOps and for the IT industry in general. You define how far ahead to look. You can talk about what happens tomorrow after we go out of the lockdown, or you can talk about what happens in one year from now, in five years, in 10, 20 years, maybe in the far, far future when there are no more humans and only machines on Earth go wild. And your timer is starting now.

[B]

So this is not the easiest question in the world, is it? I'm. Well, one thing I hope, really for the future is that we get away from DevOps as a role. The promise of DevOps for me seems to have been squashed by a bunch of consultants coming in, trying to make a place for themselves. And what excited me about it when it started was about the company and teams kind of coalescing around a piece of value that's given to customers, rather than just looking at new frameworks and new ways of pushing things out faster. So you can kind of see bits of it where people are adding extra bits to DevOps. You have DevSecOps, you got SEC DevOps. But really what I'd like to see is the industry moving around value. So cross functional teams, people working sanely to produce small bits of value quickly and safely to customers and having those customers have their real problem solved as soon as possible. But then that's what gets me going.

[A]

Sorry, isn't that what's happening today?

[B]

It is in places, but it's not evenly distributed. There's a lot of people doing DevOps and doing Agile, but what they're really doing is the same old stuff. They're finding requirements, putting them into spreadsheets and getting people to develop them, throwing them over the wall. They're doing it quicker, but they're not actually producing real value because you still have angry customers who don't get the things that they want or get the real problem solved.

[A]

So do you believe we're heading toward DevOps Haven, where DevOps is evenly distributed everywhere and all the customers are happy? Would that be a correct definition?

[B]

Gradually. Not quick enough for me. Okay,

[A]

and what happens after DevOps is evenly distributed everywhere? Do we stay in this status quo of omnipresent DevOps or there's something that happens next.

[B]

Well, DevOps goes away is the thing. It's not DevOps anymore. This is software development.

[A]

Okay.

[B]

DevOps is kind of the state we have to move through to get to the place where software development is delivering small pieces of value based on what real people want to do. Okay.

[A]

Okay. So basically what you see in the future is that sanity that you so much crave for.

[B]

Yes, definitely. And you'll still have the consultants going in, trying to do their weird and wacky things, and I'll bet you'll get a reaction maybe five years where the waterfall companies are coming back saying, this is how you really do software safely.

[A]

Well, being a consultant, I feel, you know, a little bit troubled, you saying consultants like there's something bad. But I also like to think that I'm a different kind of consultant, so maybe, maybe that's okay. Yeah, there are those bad consultants.

[B]

I show my own bias here. Right.

[A]

Okay, Damien, and you're on time again. And let me tell you how much. Oh, we still have four more minutes for some unexpected questions. So one question for you. What the hell is adaptive behavioral analytics? But to make it short, the key

[B]

difference here between the sorts of models you'll see with traditional AI and the sorts of models that you get, adaptive biometrics, is that the models themselves retrain based on data that comes in. So when you make a model, you have to train it up on a set of data, and those data will determine how it classifies things as either fraud or not fraud. The key difference, the adaptiveness of these models and how the software works, is that it's not just that initial training set that works as data, and as data patterns change because fraud changes daily as people try and find new ways of stealing money, though the models themselves can adapt and reuse those new data to retrain, catch different types of fraud as they come in.

[A]

Okay, so adaptiveness is basically about retraining each time. Okay. Okay, sounds good. Okay, what's your favorite programming language ever?

[B]

Still love PowerShell. If I ever get to try, get to program again or do anything, I usually use PowerShell mainly because it's what I'm comfortable with and it just makes it really easy. Okay.

[A]

And the most hated one.

[B]

Love, hate relationship, really, but it has to be Bash.

[A]

Okay.

[B]

Okay. Bash gets things done, but it's also extremely ugly and can explode if you look at it the wrong way. Okay.

[A]

Yeah, that's the reason to love something also because, you know, this Feeling of danger.

[B]

A little bit co dependent.

[A]

Right, okay. Have you ever had a great manager?

[B]

And what I've been lucky enough to. I've been lucky enough to have several great managers. The key thing with them has always been that it's never felt like a one sided relationship. They've always been someone who you sit eye to eye with. So it's something that John Lennon and Paul McCartney always talked about their relationship, that they were always, they were never good enough on their own. They always were better when they were eye to eye. And that's how I feel about the managers I've had who have been very good, that they manage by showing you where to go and you don't even know you're being managed at that point.

[A]

So in your notes for the meeting you said you moved to uk, so why did you move to uk?

[B]

So Ireland has always been a bit of a. Had a love hate relationship again with success and money. Weirdly in the 90s when I started working, there was like a little mini boom that kicked off and everything was going great. But then we had a bit of a recession about the start of the century and at that time struggling to find a job. This one came up in the uk. I thought let's give it a shot for a couple of months, see how it goes. And that was 18 years ago. Now

[A]

what makes a great engineering team great?

[B]

The big one is communication. So being able to talk to each other. Engineering is a team sport, which, you know, engineers aren't great at.

[A]

And Damien, we're out of time, so thanks a lot and yeah, I'll. I'll get in touch later on actually. We can continue talking.

[B]

Actually, actually hit the gong this time.

[A]

It was fun chatting to you. Have a nice one and let's talk again sometime.

[B]

Yeah, definitely. Bye Matt.

[A]

Short and sweet. Thank you for listening and watch out for new episodes of DevOps Shorts.