What happens when four decades of hands-on engineering experience meet today's fast-moving AI and open-source ecosystem?
In this episode of Tech Talks Daily, I sit down with Paula Paul, Founder and Distinguished Engineer at Greyshore, to explore that intersection. Paula began her tech career in the early 1980s writing code for IBM mainframes and has since become a trusted voice in enterprise modernization, cloud adoption, and open-source governance. Today, she advises organizations on how to build with confidence in an era defined by complex software supply chains, generative AI, and evolving SaaS models.
We recorded this conversation during a time when many organizations are feeling the pressure to adopt AI and modernize legacy systems, but struggle with where to begin. Paula offers grounded insights on how to break down replatforming into value-driven streams and why a "big bang" transformation approach is often more risk than reward. She also talks about the critical role of open source in business today—not just from a tooling standpoint, but in terms of responsibility, transparency, and community.
Paula's perspective is refreshingly practical. She believes AI is a natural evolution of decades of computing and storage expansion but urges companies to start with targeted experiments and thoughtful team collaboration. She also shares her experiences as a woman in tech since the 1980s, reflecting on what's changed and what hasn't and how leadership dynamics continue to shape opportunities for women in the industry.
Beyond the code, Paula draws a fascinating parallel between music and technology. As a board member of the Brookline Music School and an oboe and English horn student, she explores how musical thinking can enhance software development's rhythm, structure, and leadership.
If you're navigating modernization, curious about AI's role in development, or wondering how to make open source work for your organization—technically and strategically—this episode has something for you. How are you balancing technical ambition with long-term value? Let's continue the conversation.
[00:00:04] Technology is constantly evolving, but few have seen its transformation quite like my guest today. Her name is Paula Paul and she's been shaping the tech industry since the early days of mainframe programming in the 80s.
[00:00:19] And today she continues to lead digital transformation efforts as the founder and distinguished engineer at Grayshore. And with a career spanning four decades, she's witnessed and driven some of the most significant shifts in computing from distributed systems to cloud adoption, AI and the rise of open source technology.
[00:00:40] So today I want to hear more about her insights on AI's growing role in software development, the importance of open source awareness and why organizations maybe need to rethink modernization strategies if they are serious about remaining competitive.
[00:00:57] So if you're curious about where AI fits into your software development, how open source is shaping business strategy or just how to successfully modernize tech without getting stuck in a 10 year replatforming project, you're going to enjoy this one. But enough from me, it's time for me to introduce you personally to Paula Paul. So a massive warm welcome to the show. Can you tell everyone listening a little about who you are and what you do?
[00:01:25] Sure. My name is Paula Paul. And I've always from my career, I've always been a software engineer. I started with IBM as a programmer in the early 80s, believe it or not, mainframe programmer. And I've just always stayed with technology and enjoyed it like this ride through all sorts of waves of change. Currently, I do independent consulting.
[00:01:50] So I've been taking contracts in all sorts of areas to help organizations with their technology. And you mentioned programming in the early 80s there. Well, that was one of the things that attracted me to you because you've had this incredible career spanning four decades in technology. And right now we're all getting excited about AI and the shifts that we're seeing there. But obviously, this is not our first rodeo.
[00:02:14] We've been on so many different digital disruptions and technological changes throughout our lives. So what would you say are some of the biggest shifts that you've seen? And how do they shape the way that companies approach software development today? Oh, goodness. So I always tell people, especially with the current focus on AI and the excitement, we've been asking computers to do things for us for a long time. And it's all just about our comfort with it.
[00:02:42] So I'll tell you the story of I had a job at IBM. And one of the things I did early on was help IBM with its adoption of CAD software. So they were a manufacturing company and they manufactured mainframes, you know, mechanical parts. And they had people on drafting boards. So I took mechanical drafting in high school. My father was an engineer. And so that's how I actually got into an internship. And I had to show these draftsmen on drafting boards how to use CAD.
[00:03:11] And I remember the first time I showed my father a CAD system. He was like, what witchery is this? I could never use a computer. And so, you know, the thing that scared him was when you draw a front view and a side view and a top view of a widget, you could then draw what's called the isometric view, which is an angled perspective. And with CAD, you could just draw the front side and top views and say, voila, draw the isometric view.
[00:03:39] So this has some relevance in the age of AI. So I think, you know, even past that, I've always seen change. I remember IBM coming out with two-phase commit in databases was like the bee's knees because that was somewhat of the beginning of distributed computing and distributed data.
[00:03:59] So, you know, it just hasn't stopped from there through, you know, cloud, Kubernetes, you know, orchestrated computing to blockchain to AI. So it's, you know, it just is a never-ending ride. And that is one of the things that I love about technology. Yeah, completely agree with you. And it blows my mind sometimes when I look back at it throughout my own career. And it wasn't that long ago where we didn't have mobile phones.
[00:04:29] We didn't have high-speed internet. And I remember those days vividly of that sweet sound of a US robotics modem connecting. And obviously everything's gone mobile. And we've seen so many big changes. How do you think those big changes, like the arrival of the internet and the mobile world that we all take for granted and the big change of AI now, would you put AI on the same level as those changes before?
[00:04:52] Well, I would say it's a very significant change in the realm of all the things that certainly I've seen. I think that one of the challenges is that it's a natural outcome of the availability of compute and storage. You know, all the way along through all of these changes, it's about your availability of compute storage and network. And there are a lot of concepts lumped into AI.
[00:05:20] AI. So large language models, RAG, all the, you know, generative pieces. And I think that what will have to happen is they'll have to settle into their niches. Because if you even think about, if you've seen a Pixar movie, you know, it's like computer generated animation has been with us a long time. So we didn't call it AI. It's like, it's a little different.
[00:05:46] But, you know, I do think that from a software engineering perspective, it's very cool to see AI assisted coding. I think that's going to change the way a lot of people write code. But heck, I used to code on punch cards. So, you know, I'm used to changing the mode in which I create code. Oh, absolutely love it.
[00:06:13] When I was doing a little research on you, another area I know you're passionate about is you're a strong advocate for open source technology too. So how can businesses balance the benefits of open source innovation right now with those challenges of security and software supply chain risks and all those other things that we hear about? Anything you can advise there for any business leaders listening to our conversation today? Yes, become aware.
[00:06:40] I actually still work with some organizations that say, you know, we don't do open source. And I'm like, well, you know, you have a website, correct? You're dependent on the JavaScript ecosystem. And what I think a lot of organizations still don't realize is how many dependencies they have on open source libraries. Maybe five years ago, it could have been any given software product, maybe 100 or 200. And now it's 5,000, 8,000.
[00:07:08] And so you may be depending on some software packages that are vulnerable. And it's great if you scan for those things. But I would say get a little more deeply involved in your dependencies and support. Node.js is a great example. It's running, you know, a tremendous amount of the world's internet infrastructure. And it's completely supported by a foundation, the OpenJS Foundation.
[00:07:37] I would say get involved. You can become a member for a very small fee, but that's a way to support that ecosystem, including its security posture and security practices. So get involved. 100% with you. And another saying that has become ubiquitous in recent years is that every company is now a tech company. But another thing that attracted me to you is I read online that you argue that every company is actually a SaaS company.
[00:08:05] So can you unpack what that shift means and why that's important too? Yeah. Yeah. Yeah. I think it is that we are all depending on tech now in order to deliver services to our constituents. You think of a bank, online banking, websites, mobile. Everyone who has a mobile device or a mobile app, that is software as a service. So you might think of it as, well, that's your product or whatnot.
[00:08:32] But if your main value, you know, if you have a value stream and the value you deliver to your customers is, you know, a financial plan or your bank account statements, you're delivering those things through software. And so the more you realize you're truly a software as a service company, you can start to optimize and fine tune your products the same way that other SaaS companies do.
[00:09:01] And another big topic that we hear about a lot is technical debt. And legacy modernization is a notoriously major challenge for many enterprises. One that springs to mind is the travel industry. We look at any major airline out there, much of the tech that we see now is built on top of tech going back to the 60s, 70s and 80s. But because they don't have the luxury of downtime, the same goes for the critical infrastructure as well.
[00:09:28] But what common mistakes do you see companies make when replatforming and how could they avoid them? Yeah, I think really starting with the value stream. I actually got called in on a consulting opportunity to look at an old finance system. And it was running on a tandem nonstop, which is like, oh, that's back in the day, you know. But it was certainly a rock solid system.
[00:09:58] The cost of that system was really about the maintenance, you know, the yearly maintenance that you had to pay for in case it completely went casters up. So they were like, should we replatform this? And I was like, well, it's solid. It's not going down every night. You're not, you know, running around with your hair on fire because of this. And then the cost of that is certainly one of your costs of doing business.
[00:10:23] But if you replatform this, how much more revenue is it going to drive to your company? How much better service is it going to give to your constituents and customers? So I think that's a way to approach modernization. You can't always avoid those things. A lot of those old systems have become unreliable.
[00:10:47] And then that's a problem because that's stopping you from reporting your quarterly financials, your, you know, the information that you must deliver to your stakeholders and shareholders. So I would say start with understanding your value stream and then where replatforming something that is old and perhaps unreliable is going to drive revenue or better service to your customers.
[00:11:15] And, you know, I hail from ThoughtWorks very fortunate to have had that opportunity. And I would say there's a lot of great information on the ThoughtWorks website about this. But, you know, just avoiding the big bang and just, you know, 10-year plans to replatform everything. Things change on a month-by-month basis these days and you can't afford to think about a 10-year replatforming project.
[00:11:42] And we said at the very beginning of our conversation, we talked about the arrival of AI, how it's transforming everything. And I would argue that AI is transforming software development more than many other areas. But it also introduces new complexities. So how do you think companies should approach integrating non-deterministic AI technologies into those traditionally deterministic systems? Because it could quickly get overcomplicated, can't it? Yeah.
[00:12:09] And I do think that because it's so popular and so out in the forefront of the press, everyone is starting to get the fear of missing out. So I would say it's always a good place to start in a small software project and where would I use this? That should be more of a brain exercise than having people just run off and generate a lot of code.
[00:12:35] Some of the brain exercises are getting a couple of different kinds of people in a room, a product owner, an analyst, if you can, an end user. And really talk about how we might reimagine something is always valuable. Having software engineers get some training on what this kind of a tool is is always valuable.
[00:13:00] And then at the end of the day, everything that I've worked with and toyed with requires a human being to evaluate the efficacy and the accuracy of the outcome. And maybe that's the most important lesson is taking some small steps toward this.
[00:13:21] I have worked with people that say, you know, using something like Claude helped them write a Python script in 15 minutes rather than me taking two days to Google a noodle over it. So that's real. That's a real value. So don't don't shy away from those things, but don't go overboard. Yeah. Yeah.
[00:13:42] I was speaking to someone recently and he was saying that we should be using this technology to create version one, but then it should be passed to ourselves to to create that version two. And that's where the magic happens and not to get too over reliant on it. And we have talked about how much technology has evolved over the last 40 years, how many big changes that we've seen in just about everything. But I also want to highlight here that you've been a leader in a male dominated industry for decades.
[00:14:10] So how have you seen attitudes evolve around women in technology? It seemed like things were vastly improving. Still, a lot of work need to be done. But in other areas, it might now be going backwards a little bit. But what advice would do you have for women navigating the tech world today in 2025? And how can organizations better support diverse talent too? It's a tough topic, and especially in recent times, certainly in the US.
[00:14:41] I was greatly supported at IBM and had actually, I always tell the tale that like I worked for my first manager was a person of color, a color man. And then he was the manager of the B team developers. And I really wanted to be on the A team. And the A team manager was a woman. And my second level was a woman. And probably 35% of the engineers I worked with at IBM in the 80s were women.
[00:15:10] So I do think there is a corresponding graph for the US, at least, that the number of degrees of computer science degrees awarded to women peaked in 1983 or so. Like 39% were awarded to women. And that aligned with my experience in the 80s. After that, it dropped precipitously.
[00:15:34] So the number of degrees awarded to women went down to, I think, 12%. And we're starting to climb back up. So I do think this cycle will improve. I spoke to a group at a university not too long ago.
[00:15:53] And I said, the numbers are on your side because in the US, at least over 50% of the university degrees are now awarded to women in one field, like medical residencies. Over 50% of the people entering that are now female. So I do think that numbers are there, that things will improve.
[00:16:19] It's challenging that I think the people who came into leadership roles during the decline of those degrees in the late 80s and 90s are now the leaders of large companies and whatnot. And perhaps they've never had the opportunity to work with people who aren't like them. And it's a little frightening, I think.
[00:16:44] You know, I think that it can be frightening to be a friend to a woman when you might be worried, you know, is this going to get me in some HR problem? So there are societal challenges and interrelationship challenges that I think hold women back a little. And it's not their fault.
[00:17:08] And I'm a big fan of education as a very big part of the answer. But the numbers are in our favor. And outside of the tech industry, when I was Googling you there, I also discovered that you have a deep passion for music. So I've got to ask, do you see any parallels between learning an instrument and building great software? Do any of the disciplines converge? Any similarities there? Any synergies?
[00:17:38] Oh, my gosh. If you think if you've ever had the opportunity to go to a symphony, you've got all these categories of instruments, the winds, the percussion, you know, the strings. And you've got one conductor and that conductor does not play each of those instruments. And they have to get all of those people to work together on a score to produce some lovely outcome. And it's a lot like large software projects.
[00:18:06] You have to lead through influence. You have to lead through inspiration. You have to know enough about the music that each of the sections is playing in order to make it all work together. And I take a lot of inspiration from music. And also, you know, I've always had instruments in the house. And I'm learning English horn and oboe, which is like crazy hard double reed instrument.
[00:18:36] But I love it. And it's very similar feel in the brain, if that makes any sense, to what people are calling vibe coding or being in the flow. So there are a lot of parallels in how your brain adapts to being deep in a coding problem and being deep in a Mozart piece. Oh, I absolutely love that.
[00:19:02] Well, as we have a Spotify playlist, I ask my guests to leave songs and add to that Spotify playlist. As someone that's deeply passionate about music, what song would you like me to add on your behalf to the Tech Talks Daily Spotify playlist? Oh, goodness. You know, you can't go wrong with Mozart's The Magic Flute. One of the classics.
[00:19:26] And it's just like one of those that'll get you going in the morning and it does consume your brain. So if you want a little brain rinse, I do think of those things as a little bit meditative as well. So, yeah, do a little Mozart for me if you would. Oh, sure. Well, I'll get that added straight to the Spotify playlist. And we've talked about a lot in a short amount of time today. Okay.
[00:19:49] For anyone that wants to just find out more information, dig a little bit deeper about you, your career, your work, and how you might be able to help, where would you like to point everyone listening? Oh, my LinkedIn page. You know, it is linkedin.com, whack in, whack Paul-a-Paul. And then I have Medium as well, which I do write some things under. And it's usually whack Paul-a-Paul. GitHub is also whack Paul-a-Paul. So you can't really miss me on digital media.
[00:20:19] Love it. So I will add links to that. And I would urge anyone listening, if they're looking for someone to help their organization adopt cloud-native technologies, help build modern products, platforms with both speed and efficiency, you are the go-to person. You come highly recommended. But more than anything, I've just loved chatting with you today and also allowing me to have a little brain rinse in a moment by listening to that Mozart track. But thanks more than anything for joining me today. Really appreciate your time. Thank you. It's been lovely.
[00:20:50] My guest today has given us somewhat of a masterclass in navigating technology's biggest transformations from mainframes to AI, open source ecosystems, and the future of software development. When I'm reflecting on that conversation today, a few standout takeaways are that AI is here to assist, not replace. And modernization should be a strategic value-driven process, not an endless overhaul.
[00:21:18] And companies need to maybe recognize their dependence on open source tech before it becomes a security risk. So huge thanks to Paula for sharing her wisdom. But now I want to hear from you. How is AI reshaping the way you build and think about software? Do you have a story that you'd like to share? Please connect with me on LinkedIn, Instagram, and X, just at Neil C. Hughes. Let me know your thoughts. But that is it for today. Time for me to go down.
[00:21:47] It's time for me to rinse my brain and listen to a little bit of Mozart, which I'll be adding to that Spotify playlist, before coming back refreshed, bright and early tomorrow morning. Speak with you all then. Bye for now. Bye for now. Bye for now. Bye for now.