Andres Naranjo is the CEO of Software Toolbox – Full-stack Industrial AI.

What is your leadership philosophy?

1. Tell people where we’re going, not what to do. I do a great deal of “writing the end state.” I write what our vision is in five years, what our mission is this year, and what we should be doing 100 days out (pre-mortems, etc.). Then I give people the tools and resources they need, and I let them get on with it.

2. Learn by example. Yes, you read that correctly. We're in an extraordinary time where, because of AI, the price of intelligence has fallen dramatically. Now, it is our ability to learn new things and ask the right questions that matters. I really focus on learning new things and showing people that it can be done, that the thing that looks complex can be made straightforward.

3. It’s all about apprenticeship. Talent is the most important component in any organization: finding and developing the right talent. But developing talent is about apprenticeship. Bringing people to hard meetings, going to the field with them, working on a document shoulder to shoulder. I have little time for sitting in training rooms. I believe in passing things down live and allowing people to absorb.

4. Don’t lose Founder Mode. There is a famous — or infamous, depending on whom you ask — memo by Paul Graham, the founder of Y Combinator, about Founder mode. The core thesis is that founders need to care about / be in the details. I look at customer support tickets, at how people are using our AI products, such as jonjon.ai and chatuns.ai. I look at our abandoned e-commerce carts, our line-by-line expenses. I care about the details.

What is something you’ve learned recently?

In the last three years — since I moved back from Japan, went to a startup, and took over Software Toolbox, I've learned two very important lessons:

1. Friction and flow > goal seeking. We spend a good deal of our twenties and thirties goal-seeking: We have to get from point A to point B, climb the ladder, make a financial goal, etc. After I read "The Surrender Experiment” from Mickey Singer, I learned about “Friction and Flow” Pulling on lines where things are working, and not trying to obsess over making something that is not working work. My life has been markedly better ever since.

2. Entrepreneurs "will markets" into existence. Consultants evaluate existing demand in a way. I spent a long time at McKinsey looking at known markets that have revenue pools, profit pools, margin structures, etc., but all those markets existed. The advice was always about: how do you capture market share? Be better than competitor A?, but the brave people are in Silicon Valley, and they're the entrepreneurs willing entire markets into existence. I am working to bring Industrial AI into being; it scarcely exists today. It falls to my team and me.

What advice would you give future CEOs?

Over my 13-year career at McKinsey, I had a chance to serve at least seven Fortune 100 CEOs. They were all unique. I also had an opportunity to serve a handful of Japanese CEOs, who were also unique.

Observation is your superpower. Markets are always changing. Preferences are always changing. Your ability to go out to the field, talk to customers, and observe what is really happening and overlay that on your experience is your superpower. Don't wait for people to bring facts in carefully packaged documents to you. Go and observe.
It's shareholders and stakeholders. The US is a shareholder economy set to max. I spent two years living in Japan, which is a stakeholder economy set to max. Somewhere in between is the answer. Deliver returns, but take care of the people.
Make decisions…fast. 80/20. Nothing is more frustrating than too many facts or stories to prove a point, and decisions not being made. Make decisions quickly. If not, bring things to a vote.
Set the rails on how you run your business. Weeks, months, and quarters need to have a common cadence: meeting early in the week for important topics and late in the week for administrative topics; predictable quarters with board meetings, investor updates; and predictable parts of the year, such as when you budget, when you reset your strategy, etc. Set the rails and stick to them.
Everything is a normal distribution, an observer of the roller coaster, not a screaming rider. There will be average days, extreme success days, and extreme failure days. Just know that there are ups and downs. Stay an observer of the knowing that after a down day, there will be an up day, and after an up day, there will be a down day. Don't ride the roller coaster. Just know that it's happening.

What is something you’re proud of?

My Operating Model, which gives me the ability to be there for my family.

In my time at McKinsey, I developed a working model that was 95% efficiency Monday through Thursday, wrap up the day Friday by noon, and then connect with people. I've managed to keep that. I'm rather good at compartmentalizing. With AI, I now run at what I reckon is 120% efficiency during the week, executing across many fronts. On Friday, we have our all-hands at 11 a.m., and after lunch, all I'm doing is connecting with people one-on-one — whether that be acquaintances, people in the organization, etc. — and then I'm trying to make it out to see my son.

Unless it’s critical, I avoid work on the weekend, so I have family quality time, workout time, and I'm happy to say that I've had a lot of opportunities to be here for the family.

What’s something everyone should be reading, listening to, or watching right now?

Without a doubt, everyone should be taking some type of AI course at deeplearning.ai. They are taught by Andrew Ng, who is one of the godfathers of AI. You can get started with AI for Everyone, move on to Generative AI, Agentic AI, Coding with Claude Code, etc. We are all impacted by AI. It is important to know what the technologies are and what they do.

Beyond that, I would say the operating verb right now is experimenting. Everybody should have Claude code on their desktop or Codex from OpenAI (same thing) and be thinking about building an application. Whether you're building an application to manage your home, to manage your schedule, for your small business, your small practice, anything. There are hundreds of examples out there of people building amazing things on Claude code. The cost of building software has dropped to $20, and we're in the fast fashion or "mom and pop" era of building software. There is no need to be afraid; everything is straightforward to download, set up, and use.