Software development is undergoing rapid change as AI, DevOps, and data science reshape how teams build and scale products. In this interview, Particle41 CEO and technical co-founder Ben Johnson explains what modern software teams must do to stay competitive.

He shares practical insights on boosting developer productivity with AI, building reliable systems through infrastructure-as-code, and adopting modern data architectures that move beyond simple dashboards.

Ben also discusses how to lead remote teams effectively and apply OKRs in both work and personal life to stay aligned on what truly matters. If you’re involved in building software or leading technical teams, this conversation offers clear, actionable strategies for thriving in the AI-driven era.

Check out the full series of “Career Sessions, Career Lessons” podcasts here or visit pathwise.io/podcast/.

—

Watch the episode here

Listen to the podcast here

SaaS Software, DevOps, Data Science, And AI With Ben Johnson

Founder & CEO Of Particle 41

I’m J.R. Lowry. This is Career Sessions, Career Lessons, which is brought to you by PathWise.io. If you’re ready to take control of your career like thousands of others already have, join the PathWise community. My guest is Ben Johnson, a serial technical cofounder with a track record of success and hands-on open source programming experience. Ben is the CEO and Founder of Particle41, a development firm that aims to help companies accelerate their initiatives through software development, DevOps, and data science.

In our discussion, we’re going to be talking about Ben’s work leading a SaaS company, his thoughts on DevOps and data science, and how we’re measuring productivity in software development, particularly with the onset of AI and a little bit about his own personal productivity hacks and some of the things that Particle41 is using as well. Let’s dive in.

—

Ben, welcome. Thank you for doing the show with me.

Happy to be here.

I’m looking forward to getting to know you a bit here in what you’re up to. Let’s start with that. Give us a quick background on you and what you’re up to professionally.

I’m a serial entrepreneur usually from the technical cofounder seat. I built a company in travel, travel media and finance media. Back in 2015, a buddy and I started a company called Legalinc and sold it to LegalZoom in 2018. I created Particle41 back in 2014 out of that technical cofounding and those opportunities that I had. A lot of folks would refer business to myself and my partner. My business partner, who’s an Indian national, was part of that first travel company that I built. My head of design was the designer at Legalinc. We just amassed this, we called it from the boardroom to the code review. The service that handles. We want to be the world class tech partner for mid-market company companies.

How did you ultimately decide to focus on building SaaS companies?

SaaS Entrepreneurship & Motivation

When you’re a software engineer, you want people to use your software and that’s what SaaS is all about. It’s adding real world value with the software that you write. After doing that a few times and seeing the delight of the users that were gaining efficiency and effectiveness in their business with technology. That’s just what I found I want to do. This will sound grandiose and I don’t mean it to be that way. I also think when you’re creating complex systems, you’re imitating God in a way. He designed our bodies that are very complex systems. He orders chaos and when you build software, you get to imitate that in a way.

That architecture part has always been personally appealing for me. I used to do development many years ago. I miss it sometimes because you do get a real sense of satisfaction from seeing the transition from an idea in your mind to something which performs a business function or some other function depending on what your software’s for. I used to love doing that with too short a part of my career in the scheme of things.

I call it from agony to absolution. “Why doesn’t this work?” “I got it.” That agony to the absolutions cycle is very satisfying, I agree.

I don’t know about you. I used to probably find some of the solutions to my toughest coating issues while I was sleeping. I would wake up in the morning and I’m like, “Now, I know how to do it.” It was just the craziest thing. It happened again and again.

I experienced that or in the shower just when you are in more of a beta state and the a-ha moment would come to you. There’s some cool abstractions that have come from that like, “Rather than writing all this code, why don’t I write the abstraction for this?” It’s less good but more powerful.

With AI tools that you almost get to operate more, that abstraction layer and what a lot of the detail is spit out for you.

I agree. It’s changing. At our firm, we don’t have a spouse to vibe code because that’s this idea of asking for something big. There’s some corners being cut with that vibe coding process but more of an aim small, miss small. AI for us as an intelligent autocomplete or a pair program that if we stay in the driver’s seat, we get a lot of power and potential from that. Whereas, vibe coding to me, is asking for larger solutions where a lot of the design and ideation is coming from the hallucination rather than from the intention. That can be dangerous. You have to be willing to walk away, can and start over your vibe coded pieces or your vibe coded ideals, but it is proving concepts and all kinds of cool things.

Defining & Implementing DevOps

Let’s roll with this idea of how software development is changing a bit more. Talk about DevOps and just how DevOps has altered the way that software is developed and run.

It’s more of a methodology or a movement if you will to realize. The most relatable example is when you get a new PC or a new Mac. You go through a process of reconfiguring it and setting it up again. A lot of mouse clicks and browser searches. As that was happening in the IT market, folks started to say, “There must be a better way.” They started writing code that symbolizes, that codifies their architecture and their infrastructure. Rather than just having a deployment pipeline for your software, you also now have a deployment pipeline for your infrastructure.

Rather than going and mouse clicking around and creating infrastructure in a UI, you represent that infrastructure with code which then makes it highly repeatable. In DevOps, we use a metaphor of, “We now treat our infrastructure like cattle rather than like a pet.” This is huge. When you have an upgrade, you just change the code and redeploy the upgrade. We had clients that had invested in infrastructure pipelines before the crowd strike incident. We had them up and running and just a matter of minutes because it was just simply like, “Let’s go back to how to fix this.”

Software Development: In DevOps, we use the metaphor that we treat our infrastructure like cattle, not pets. This is a big shift. When you need an upgrade, you simply change the code and redeploy.

We wrote the code and deployed it. Everything was fixed. Whereas people who hadn’t invested in modern DevOps, they were trying to figure out, “How do I even log into the computer to change the file, to repair the issue?” That’s why many firms or many companies were down for a long time because they hadn’t invested in that repeatable process to manage that infrastructure.

If you’re making this transition into that construct right into using the DevOps mindset and model in the way that you run your company. What do you have to do to make the shift?

You do have to make an investment in the infrastructure as code. You have to build the infrastructure pipeline. Think of that like the same level of effort as building a mobile application for your business or building another software tool for your business. When you do that, you need to get clear as to what infrastructure do our applications need. There is an upfront investment but then there’s so much less mystery after that. You’re defining the infrastructure and the different application types that you’ve chosen to support.

It takes a little bit of planning and a little bit of elbow grease to write that infrastructure as code but then you emerge into this new level of maturity with your infrastructure where there’s no more mystery. The deployment process for both infrastructure and codes is gone. Businesses that are hearing things from their Tech Team like, “We need you to tell the customers that we’re going to have this long-scheduled outage.” They’re just hearing this frequently. We’re going to have to have scheduled outages and maybe they’re having unplanned outages quite frequently. That’s a good signal.

The KPI for DevOps is speed to recovery. We all know that things are going to happen or things are going to go wrong. Usually, a lemony sniffs a series of unfortunate events but your time to recover is what DevOps accelerates greatly. Something unforeseen happened, but we recovered in fifteen minutes. That’s the outcome that we see from a DevOps mature organization.

What are some of the mistakes that people make when they try and make the jump into that construct?

They don’t engage the developers enough. The main mistake is they think of DevOps as a role. It becomes the new cloud IT. Now, instead of people to go touch the user interface and do management by configuration. They now have this new role that’s management by code. They haven’t shifted much. They’re just managing their infrastructure in a different way. What we do is we say we’re doing our job well when we get fired from our success. The meaning here is we want to create an infrastructure that’s inclusive of the developers and that they know how to influence it by changing that code and submitting those change requests.

They’re a part of the process, where people make the mistake is they just shift from a legacy IT function, which is the more manual change management to now a code-based change management. They’re just moving lumber. The most successful transformations are the ones where the developers can get what they need in a safe but repeatable way. The DevOps personnel is an architectural advisor and they write some infrastructure code and help accelerate that process. Once they’ve written all that and the developers are self-serving and part of the process, then they become great architectural advisors. They’re not just a cost center, but they’re part of the growth of the company.

AI’s Impact On Productivity & Agents

How are AI and other trends affecting where this goes over the next 5 to 10 years?

I already see about a 40% productivity boost from AI assisted programming and AI assisted pipeline development, both in data pipeline and infrastructure pipelines. That 40% is felt in the beginning where everything’s like greenfield and how do I get all this code written and AI accelerates all those fingers on keyboard clacking and typing. If you think about it in a way, if the market just got 40% more efficient, then that’s why we see impact and number of total jobs and all of that.

However, on the external side, everybody needs an agent in their business now. Everybody wants that human-like experience with the machine. Your brand will need a human-like experience to go along with the brand and assist the customer. There’s a huge opportunity that puts an extreme emphasis on your data as a business and how your data is organized, your standard operating procedures, and your policies on how you serve customers.

If you’re not clear on those, if you’re hustling a little bit in those areas, where sometimes we respond to the customer this way. Sometimes we respond to the customer in this other way. That ambiguity makes it hard to leverage AI. You’ll have to reduce that ambiguity to implement agents that will bring your business into the mainstream. I heard a stat. There’s over 200,000 applications still running in AS/400. There’s still a lot of commerce.

A lot of businesses that are heavily dependent on legacy infrastructure that doesn’t make data highly available. AI enablement is a new wave that everyone is going to have to get with or die. There’s an amazing opportunity to help folks make that transition. It’s another wave of digital transformation that emphasizes your data and your customer policy in a way that we’ve never seen before.

Software Development: AI enablement is a new wave that everyone will need to adopt to stay competitive.

Data Science Role & AI Transformation Maturity

Talk a little bit more about how the role of a data scientist changes with some of these things that are going on.

Data science is separated now. A lot of my background is in online advertising or online media. We’re talking about AI and ML. What we meant back then was predictive analytics like, based on the past data, what is the future likely to look like? Now, we’ve shifted to AI as this real human-like conversation. Both things are true in the world of data science where we need to do data engineering, which is the data pipelines that make all the companies data available and in a centralized place so that the AI can converse about that data and turn that data into the human-like conversation.

The role of data sciences is now split into agentic solution development. How are we going to apply the use cases of the business, the conversations and the context of the business to the data? Data engineering is all about data availability, enrichment, and making sure that the data is coming together in an accurate way. It’s not any more about a BI dashboard. It’s about asking a question and receiving an answer, but that doesn’t mean we don’t need data dashboards. That doesn’t mean we don’t need analysis for the past. It doesn’t even mean we don’t need predictive analytics. We still need all of those things. What’s changed is now we want to have a conversation about it.

With AI on the other side of it. That ends up meaning that a lot of what you might have had to go to a data scientist to help you massage is becoming easier to self-serve. Is that a fair way of describing it?

I feel like it’s a different interface. All of the muscle behind it is still needed. All of the same tool sets and data availability. I often say, AI is not magic. It just feels that way. It’s a new interface to access the data of the company or the policies of the company or the context of the business. This is a disagreement or a contention that I’m seeing in a lot of businesses. They just want more of the AI. More of that conversational awareness of the business context. They’re all looking at YouTube and seeing the news that AI is just making everything easier.

It’s a level of maturity with data. I’ve done data engineering. I have all of the hard facts. Now, I’m applying to the agent to understand those hard facts. Another way of describing this is how AI transformation is taking place within a business. In data engineering, in data science, we have this medallion architecture. Think bronze, silver or gold. We bring data into bronze. We analyze it. We get value out of it but it’s a single source value when it’s in bronze. You’re bringing in your IVR, your call center data and you’re like, “Now I know how many calls hold times.” You can report off of those single sources and get a lot of value.

In AI transformation, the synonym to that is, “I’ve given chat GPT to all my employees. I’ve given them a couple of prompts and now they’re so much more efficient in writing emails.” The user is doing the orchestration. They are maybe pulling in a little bit of data or giving some context and then getting an output then taking that output into an email or into a presentation or a spreadsheet. That part of the transformation is interesting because you probably are getting a ton of value but it’s not combined or it’s not in a workflow.

The next stage of AI transformation is what I would call silver where you’re thinking about the workflow. How can I get data from these different points, give it to the AI, and think of different use cases and scenarios. There is still a lot of traditional programming happening in that workflow, BPO business process optimization stage, and then you get to gold, which is the agentic. I know all the use cases that I can.

I’ve delegated first tier customer support to the agent. Now, I have a chatbot on my website. Folks are going through this transformation from bronze, silver to gold and they’re underestimating the work that it takes to go through all of those stages but it is inevitable that you would start with just some prompt engineering and prompt enablement then progressed to specific scenarios that make sense to mature all the way to the agent level.

AI to your point, it isn’t magic. It may feel like magic but it’s not going to solve all of your underlying data issues in a way. I remember thinking this when I was hearing somebody talk about what one of my former companies was doing with AI. This goes back probably a few years, Ben. Even before we had all heard about ChatGPT.

He was talking about what they were doing and I just thought, “This just puts your data issues on steroids.” Any shortcomings you have in your underlying data, this tool is just wanting to consume massive amounts of it. If you don’t have the ability to get that data and to have it be clear, complete, and accurate and all of that, then the AI tool can only do so much magic solving of that.

That’s right.

What are some of the other things that you see forthcoming in the data space as it fits into the broader landscape of SaaS software development and the businesses that you’re involved in?

We’re seeing businesses where tech or SaaS is not their core competency. We’re seeing manufacturing firms be able to do things like monitor waste and needle moving things at a much lower cost and effort where we used to build the data and understand all the different data types. It used to be a small data team. That data teams execute more faster for a greater result than ever before. The acceleration of your competitors that are adopting it, like the fear of falling behind should be a greater fear than it has ever been.

That’s the fear we want to address with our firm. It’s how we keep you from falling behind, how we enable you with some of these latest and greatest tools. The cloud providers are eager to compete with one another and assist you in getting there. There’s funding available with many of the cloud providers. There are all kinds of opportunities to not fall behind and we are just enjoying assisting folks in that journey. Even in industries that have been poorly tech enabled in the past are finding the value from what data and AI is able to do for that.

Which ones do you think are particularly lagging?

Manufacturing is an interesting one because if you can make the widget and you’re not cost pressed. There’s a little niche. Property management space is an interesting one that we’re seeing traction in. Also, places where there’s an incumbent that’s not taking the time to innovate. If you look at the best innovators, they risk disrupting themselves. If you think of the Netflix Blockbuster example. When you see a company that’s risking its own disruption. That’s a very cool thing.

It’s rare. This is what made Clayton Christensen’s book The Innovator’s Dilemma such a huge hit. Industry after industry, it’s very common that the incumbents don’t want to disrupt themselves. The time frames that disruption usually entails ends up having a trough period that runs longer than most of the executive teams. It feels like they can get their investors or their own careers to stomach. They don’t want to make the investment period or they don’t want to go through the investment period. They don’t do it and you have the Blockbuster Netflix situations. Where the new guy takes out the old guy because they were willing to disrupt the industry even if the incumbents weren’t.

AI Risks, Regulation, & Societal Impact

I hope that there’s an innovation in the more heavily regulated spaces. Finance is trading a stock. It isn’t the best experience. They will be something. A lot of folks want to auto trade or leverage different algorithms for trading. That will be an interesting one to see how the market adjusts to that because if we all went and got a trading bot, how would that change the market? Healthcare is an interesting space too because when is it appropriate to ask an AI for your health care plan rather than ask a doctor? How will that progress? How will we deal with that? There’s so much innovation potential in heavier regulated spaces where AI is the most scary to people.

Where things can go wrong in a way that is very damaging. It could be in the finance sector where algorithms go off the rails and take the market down. That’s already happened. It could be in the medical industry where AI ends up canceling somebody to commit suicide. It’s already happened. There are some pretty bad stories out there where these tools have gone off the rails. That’s why there’s hesitancy and industries where the consequences potentially can be big. It’s a little bit more amusing when the AI goes off the rails and offers you a $50 ticket to Europe on an airline. It’s a little bit of money. It’s not necessarily a life altering situation.

Whenever I go to a restaurant, I take a picture of the menu and I have a prompt that describes my health goals. I usually get 3 to 5 recommendations from the menu that includes substitutions on how I might want to navigate the menu to be the most healthy. That’s super convenient and I’m not paying for a natural path to give me the same coaching or I’m not paying a health coach to do that. That’s both valuable and interesting.

The scary part of that is if you’re a young man you can either choose to enter the dating life as a teenager and start going through all the fear of rejection or you can go get an AI girlfriend. Those two paths are quite possible. The concern of that, as a parent. That’s where maybe not all AI solutions are going to be good for us as a society because I can ask my wife to be more affirming. That conversation generally teaches me a lot that is not necessarily very affirming but if I ask an AI bot to be more affirming then I just get instinct compliant. That is how we start to use it in artificial relationships and how we personify it in our lives. That’s going to take an interesting level of moral responsibility.

Software Development: Not all AI solutions will be good for society. Navigating this will require a new level of moral responsibility.

I don’t know if you watch South Park. There was an episode in part focused on the plan during which ChatGPT is poked fund for where the wife starts answering her husband like ChatGPT would. He’s like, “This is a great conversation. Thank you so much.”

You can ask the wrong question and you don’t get like, “That question is flawed. It’s not telling you when you’re asking the wrong question.” How much responsibility we give it will be a concern because then you could start asking all kinds of wrong questions. As a society, I feel like we’re very responsible for information as it is. We’re pretty irresponsible with the information age in the same way we were irresponsible with the industrial age with a lot of pollution.

One of my great grandmother’s household remedies for skin problems was to spray Borax or whip up a very chemical solution for certain things. We started to learn that we need to be environmentally responsible and apply some caution to industrialization while we’re still in that maturing process with information. We have AI and the potential for AGI to come in just hope and pray that we get a little more responsible a little quicker.

You can be both. You talked earlier about some of the efficiency boosts that you’re seeing with using AI to support the development process. At this point, being involved in leading a company. What does an efficient technology team look like to you? How would you define efficiency in the context of a technology team?

Defining An Efficient Technology Team

You want to be represented from multiple roles. You do want the software developer, the DevOps guy, and the data guy. The roles need to be covered and what’s going to get stress tested a little bit is the idea of agility. We are going to plan more and design more upfront. It’s a lot cheaper to change the Figma design than it is to change active code. That still continues to be the case. We’re going to go back to a little more planning up front because the software factory process won’t be as human driven. It will be some humans with some agents.

I do think that a small team is going to be able to do more with less, but I do still think that those separate roles on the team need to be represented so that you can have that healthy debate within a team on this separation of concerns and the right architecture. I do think it’s going to be a challenge for the up-and-coming talent. The folks that don’t necessarily know how to design those systems, but I’m sure as all young people do. They come up with new ways and new means. They figure out how to become relevant. I’m excited to see what people entering the field come up with. Stay relevant in a rapidly changing space.

Our teams are able to do more faster and we’re ending projects more frequently because we’ve successfully accomplished the mission. That efficiency in our model, we have more of a managed capacity model. That savings from the AI efficiency is being passed straight onto the customer, even at some challenge to our own business. The barrier of entry to certain solutions is rapidly reducing.

Does the efficiency automatically get taken as savings or does it give you more latitude to experiment more, be more creative, or more innovative?

Let’s say you wanted to build a software solution. You’re going to be able to do more with less. If there is competition in the space, you can equalize your feature set rapidly to match your competition and then apply on top of it your differentiator. You can come up to the level of competition. Whereas before, that required a significant investment to get to the same place as where your competitors are. That me too cycle that’s entering a space might go through, that part is accelerated and then you figure out, “How do I test the differentiator? How do I make this different or better than the competitors? How do I apply better customer service to this?” You can exploit your competitor’s weaknesses much quicker.

How do you measure whether a team is truly efficient other than velocity metrics?

Importance Of Culture & Trust In A Remote Environment

Velocity certainly matters but also, for me, how they participate in the practice, and how they participate in the abstraction. Do they work on the business, not just in the business? That sounds subjective but we have some clear things that we want to see as we’re evolving the practice and their participation towards that. Also, how they connect with one another. We’re a 100% remote organization. How do they show up for each other, how do they take initiative with some of the culture or are they fairly transactional about their role? You pay me, I do my work and I’m out. Those cultural things are relevant in a world where the time to execute a particular mission is reducing.

You’ve obviously learned a lot about how to be successful in a fully remote environment. What advice would you give people who are struggling with getting people to bring that level of commitment while operating in a hybrid or remote environment?

There’s a trust button verify. What’s important for us is visibility. We have a bot that reminds everybody to check in daily and that becomes very important. Have you decided as a business to track hours? If people are delaying their hour tracking, that’s a bad sign in a highly remote environment but we don’t log hours. We just report the day.

Everybody has to be measured by their results. Not necessarily they’re working hours or if you think they’re online. It’s more like that they’re talking about the work actively in asynchronous communication. They’re good at asynchronous communication and there’s just no question. As the professional, it’s like are you checking in daily? Are you sharing what you’re working on? Are you sharing what you learned? Encouraging everybody to keep up with that and do that is part of the role. That’s what we’re seeing.

I like the way you put it early or about, are you working on the business and not just in the business. It implies the level of commitment to collective success and not just, “I’m just doing my thing and I’m clocking out. Have a good day.”

It’s important. It’s just very noticeable for us. If you’re out of sight, out of mind, with remote, that’s a bigger issue than the quiet person in the corner of a physical office.

Very true. You talked about that bot that you created to have people check in each day. What are some of the other tools and systems that you use to run your own company?

Using AI In Internal Business Processes (Recruiting & Reviews)

The AI enabled the recruiting process. I know there’s a little bit of critique on this but we do score inbound applications and we use an assessment tool that has some AI in it. We want to make sure that the resume matches the job description. We understand why it’s being scored low or high. For example, SAP Engineer with twenty years of experience is not appropriate for one of our data engineering positions.

They could be a data engineer but it’s just a different data engineer. They get scored low even though they have high years of experience. We see that in our scoring methodology. We love to talk to all the people who pass the assessment and assessment is managing a cheater score like, are they in there? That has been cool because we feel very effective and selecting our folks, then they go through a rigorous in-person panel interview and all that.

Have you tried using AI for the interview itself?

No, in fact that’s where we’re wanting to make sure. Our interview questions have pivoted to more stories like you need to tell us a story. We’ve found that when people are using AI to give the interview, they’ll be very academic in all their answers. Even when we’re asking for a story. It’s become far too common where folks have ChatGPT running.

It’s listening to the call and they are reading the interview. If our questions were academic, we wouldn’t be able to tell. If we ask about experiences, then we know that the AI cannot speak to experiences. We’ve shifted. After we’ve done the AI-driven assessment and we’ve scored, now we want to get to know the person and make sure that the person we’re interviewing is the person who’s going to do the work.

I can’t imagine personally the idea of turning interviewing over to AI. I thought it was funny when the companies were complaining that the candidates were using AI to respond. It’s like, this is a very one-sided thing. It’s a bit like the students getting mad that the teachers can use AI, but the students aren’t allowed to use AI. What are some of the other ways that you guys have implemented interesting tools or systems to help you run your business?

We’ve created our own system for yield and our own way of seeing resources. We do like an EOS light thing, so we’re running a scorecard in KPIs and love helping people with all that. There’s not a substitute for just good old fashioned like having an operating system, a strategy quarterly goals. We’re not using anything fancy there. We’re using a spreadsheet. You don’t need a fancy tool for everything because it doesn’t replace the need to manage it.

I also like 15/5 for employee engagement. If the leaders in the business are using the one-on-one feature and they’re meeting with their folks one-on-one. When you go to write the performance review, it uses AI to leverage the one-on-one content for the performance reviews. That twice a year process that everyone’s belabored by to get their performance reviews of their team is accelerated by saying, “If you do the thing regularly, keep it light but keep it recorded. It’s going to have a payoff for you when you need it most.

That was interesting. Having a trivial mechanism that’s AI-driven is interesting. Surprisingly, Read AI is doing a good job with this. Read AI is taking the transcript of the calls but it’s also connecting to email and documents. It’s starting to be a central point where we’re going for information about a client or just the level of communication and some follow-up on communication. Read AI is becoming a frequent first stop for, “What’s the next step? How should I prepare for this meeting? What are my action items?” It’s surprised us in terms of how it’s pulling together all the corporate information.

Personal Goal-Setting & Intentionality (OKRs)

We’ve got to be recording this conversation, so it will go into the archive. On a more personal level, Ben, how do you keep up with all of it? Running any company in a growth mode, starting from scratch, and building from there is always hard. How do you stay on top of it without burning yourself out?

I am relying on the practice leaders to spend that time and bubble up what’s working and what’s not working. I have my own hobbies and interests. I have a home lab here at the house. I’m using AI to teach me how to do certain things. Being a participant in the learning process is helpful and then a book a month. The ONE Thing is a great book. I’m trying to digest a book every month and internalize it like making it part of who I am.

A few years ago, as a company, we read Measure What Matters, the quintessential OKR book and I put my teenage sons at the time. Now, they’re in their twenties but I put them on OKRs. It was transformative for them, like thinking about, “What do I want to be? How do I get there?” One of my sons said, “I want to be a professional gamer.” “How are you going to manage your progress?” He’s like, “I have to play more competitively than recreationally.” “You have to play your hours per week of competitive play that’s organized into a league of some sort.”

They went and signed up for a league and started to feel the pain of organizing that and then realizing that there was like minor-minor league, minor leagues, major league and pro and like, “Am I going to do all of that?” All those guys stream. That’s how they afford their hobby. Are you going to start streaming? When the cost was measured, the goal changed.

My oldest son graduated magna cum laude from DBU in 2024 and is now enrolled in his doctorate program. It was that goal-based thinking, that intention of, “What do I want my profession to be? How am I going to progress towards it?” That made him the most successful just thinking, “I won’t be able to get into my doctorate program if I don’t have a 3.8 or better. I do have to study.” That quantifying was just what bringing in OKRs into the family. This idea of quantifying success was transformative for all of us.

Did you try them on your spouse?

I got less participation there but we have an OKR now about the number of double dates. We have a vision to impact couples. We want there to be more successful marriages in the world. We want to go on a double date once a month. That’s an OKR we have, and it’s been great. We’ve gotten to be intentional on those double dates about asking our friends like, “How is your marriage going? How is it going well or poorly? Do you guys have some goals that you’ve set for your marriage? Is there a recurring spiral that you want to talk about?” Seeing our friends open up, be more vulnerable and transparent rather than just small talk on a double date has been awesome.

Do they know what they’re in for when they agree to go out with you or do you spring it on them?

Absolutely not. They’re appreciative, though, afterwards of the realness of, “This wasn’t just another double date.” We can read the room. We know if we need to just keep it casual or whatever, but it’s great to align that way.

What’s ahead for you and Particle41?

We’re doubling down on the agentic system programming. We’ve had a thriving data practice. We want to continue to help people mature towards agentic AI and we’re enjoying that. We’re doubling down and just want to see how we can help customers do more with less.

If you could go back and give your younger self some advice about how to manage your career or maybe what the advice you’ve given your kids. What would it be?

That’s a very great question. Setting an intention rather than being opportunistic would be great. I would have become a business owner sooner and there’s some other things I would say about knowing myself like what I know about myself now. Taking some time to smell the roses and enjoying the people around you. would have made the journey a lot more enjoyable.

That’s always good advice. Do you feel like you stumbled into being an entrepreneur, or was it something that you always had an idea of doing?

I was intrapreneurial from the very beginning. There was always something about, “If I’m right with this piece of code, I know where it started.” We built the travel company and I was a young buck. I got well mentored. The CTO there was a Harvard MBA, MIT undergrad. He just tore me apart and put me back together again. He took the keyboard and mouse off of my desktop computer for a week because I wasn’t managing and leading. I was having everybody help me be more successful rather than helping everyone else be more successful.

You were working in the business.

I was working in the business rather than on the business. Some of those lessons that he taught, and we built an amazing travel platform with air, car, hotel, started directly connecting with brands and then we sold it. The buying CEO, the purchasing CEO made so many decisions based on his personal biases. Not because they were good business decisions but he had a call center in Vegas because he liked to be in Vegas. He had a call center in Atlanta because he liked to live in Atlanta.

We had a thriving call center that had bilingual, low cost of living in West Texas. It’s a place executives don’t want to hang out because there’s not that much to do and it was cost effective. To make these like biased decisions and just kick the can down the road, that impressed on me like, “I can create the best software in the world, but if there’s not the right business strategy around it.” That just made me super curious about every venture. What’s the strategy? What are we building the software for? Are we going to exit? Is this a lifestyle business for you all? What’s the strategy here?

It made me so hyper aware of that because I poured everything I had into all that travel tech. It was a lot of hard work and just to see it in basically eighteen months in bankruptcy. From 70 million in travel successful business. Travel is a very thin margin business, but it was successful to see it go right into the tank. It got me curious about how I can be a better business person.

To your point, a lot of times people get into these positions of power and they’re just so enamored with the idea that they can make the decisions. They make decisions that are gut-based or preference-based but not necessarily fact-based and objectively sound decisions. Deciding that you want to put a call center in Miami because you like visiting Miami. Not the cheapest place to set up a call center. It would be an example. That Las Vegas is probably a little bit more cost effective, but maybe not anymore. Making these kinds of decisions for the wrong reasons is basically betraying the organization when you do that as a leader and a lot of people unfortunately forgot that.

You’re making good business decisions, you know your numbers, you have your fiduciary responsibility and then you treat people like people and not tools. It’s such an easy disillusionment. I resist the word resources. They’re not resources. They are people. They are teammates. If I hear myself starting to say resources, I always change it to teammates because they are my teammates.

When that shifts, then the decision becomes for the collective good of the business and everybody’s livelihood rather than, “I want the gross margin to be 2% higher this month.” It’s like, I need to grow margin to be 2% higher this month so that we have a successful business so we can put more into sales and marketing. You have more career opportunities within the business. It all fits. We are in this together.

There are a lot of lessons in there too that you can all in part to your younger self or to your kids. Thank you, Ben. I appreciate you making time to do this and great to hear what’s going on in the software development space because it’s a rapidly changing one.

Thank you so much, J.R. I appreciate you.

—

I want to thank Ben for joining me to discuss his current work, what it’s like to build a SaaS company, thoughts on DevOps and data science, measuring efficiency and software, and some of the personal productivity hacks that he employs as well. As a reminder, this show is brought to you by PathWise.io. If you’re ready to take control of your career, join the PathWise community. You can also sign up on the website for our newsletter and follow us on LinkedIn, Facebook, YouTube, Instagram, and TikTok. Thanks.

Important Links

• Ben Johnson on LinkedIn
• Particle41
• The Innovator’s Dilemma
• Read AI
• The ONE Thing
• Measure What Matters
• PathWise on LinkedIn
• PathWise on Facebook
• PathWise on YouTube
• PathWise on Instagram
• PathWise on TikTok
• PathWise on Twitter

About Ben Johnson

Ben Johnson is a serial technical co-founder with a track record of success and hands-on open-source programming experience. His experience includes being a board-level advisor, founder, and hands-on operator. Through his 20+ years as a software developer and leader, he has gained extensive experience with remotely distributed development teams and developed an in-depth knowledge of business hacks.

Ben is the CEO & Founder of Particle41, a development firm founded by industry veterans that aims to help companies accelerate their initiatives through Software Development, DevOps, and Data Science. With a constant focus on results and ways to improve, he is having fun building highly scalable and highly secure applications.