Your Personal Tech Stack Matters More Than Your Company's

I am increasingly of the opinion that your personal tech stack matters. Not in a peripheral, nice-to-have kind of way but rather in a “this will shape the trajectory of your career” kind of way.

You’ve probably heard some version of “you won’t be replaced by AI, you’ll be replaced by someone using AI.” There’s truth in that, but I think the real story goes deeper. The professionals who will pull ahead are the ones who capture and manage the data that makes them effective, so they can layer on AI and scale their impact. I’m not talking about becoming a prompt engineer or building apps. I’m talking about being intentional about the tools you use every day and making sure those tools, and the knowledge they hold, belong to you.

I’ve argued previously that the future of human-centric AI adoption in construction is B2C, not B2B. That enterprise AI optimizes for margins and systems while individual AI builds personal leverage. This article is the practical follow-up. If you bought the “why,” here’s the “how.”

What I Mean by Tech Stack

When I say “tech stack,” I’m not talking about your company’s software suite. I’m talking about the personal ecosystem of tools you’ve assembled to make yourself a more effective professional. Or haven’t. And if the latter is true, fret not, I’ll walk you through my journey and you can hopefully pick and choose from my experiences.

For starters thought, here’s what I mean by “tech stack”.

Where do you store your contact information?

Every professional interaction you have on projects, in meetings, at conferences, is building your network. But if those connections aren’t captured in a way you can recall them when you need them, you might as well not have had them. A business card in a desk drawer or a LinkedIn connection you can’t find does nothing for you when you need to staff a team or get an introduction.

I use Clay for this. It’s built specifically for relationship management rather than traditional CRM, and it lets me maintain and actually use my network in a way that scales with my career.

Where do you record and retain meetings?

I’ve used Otter.ai, Read.ai, and Granola.ai. They all have their strengths and weaknesses, although I am currently loving Granola. The point isn’t which tool you pick. The point is that you have one, that you own it, and that the institutional knowledge from every meeting you attend is being captured in a system that travels with you.

What LLM are you building memory around?

This is the one people overlook. If you’re using ChatGPT or Claude or Gemini through your company’s enterprise license, every conversation, every piece of context you feed it, every workflow you’ve refined lives on their servers under their terms. When you leave that job, that context stays behind. Make sure the LLM you’re investing the most time in is one that will go with you.

Full transparency: I use Claude. I’m writing this article with Claude helping to ensure my run on sentences are less offensive to your ears and grammatical challenges are less pronounced. I’ve tried all the major models extensively, and Claude is the one I keep coming back to. It handles nuance well, it pushes back on my thinking when I need it to, and the memory features mean it’s building a genuine understanding of how I work over time. That said, the landscape shifts constantly. The model that’s best for you depends on how you work. The principle matters more than the brand. Own your LLM relationship.

What narrow AI tools does your specific workflow require?

General-purpose LLMs are powerful, but they don’t cover everything. I work a lot with visuals, presentations, one-pagers, pitch decks, and I invested in a Gamma license to help me create those. It’s been a big win for my workflow. Think about the specific outputs your role demands and find the tools that accelerate those. Don’t wait for your company to approve something. Invest in yourself.

Where to Start

If you’re reading this, you’ve probably already experimented with AI tools. Maybe you’ve had some good conversations with an LLM, used it to draft an email or summarize a document. But there’s a difference between dabbling and building a system. Here’s how to get more intentional about it.

Accelerate your learning by leaning on others.

I took a Masterclass on AI early last year that really helped lay a foundation for how I think about leveraging these tools. I also read (full disclosure I listened to the audiobook but that counts right?) Ethan Mollick’s book “Co-Intelligence”. The point is, you don’t have to start from scratch. Take a class, watch some focused YouTube content, read what people who are further down the road have figured out. The goal isn’t to become a technologist. It’s to build enough fluency that you can make informed decisions about what belongs in your stack.

Find a voice transcription tool that you own.

This is one of the highest-leverage moves you can make. The amount of valuable information that passes through your day in verbal form, phone calls, site walks, OAC meetings, is enormous. Capturing that systematically changes the game. Some options worth exploring: Plaud.ai, Granola.ai, Otter.ai, and Microsoft Copilot. The key is that you own the account personally when possible.

Get a personal pro license of an LLM.

Pay for it yourself. ChatGPT, Claude, Gemini, pick one. The reason this matters is portability. If your company is paying for your Copilot seat and you leave, all the context and memory you’ve built is gone. When you pay for your own license, that investment in context compounds over time and follows you wherever you go. Claude is my pick right now, but the important thing is that you’re building on a platform you control.

Develop a foundational prompt methodology.

Every thought leader on the internet has a guide you can buy to help you prompt better, but I’ve found value in keeping it simple. The LLM does best when it knows what you want, why you want it, and the perspective you want it to take as it approaches your request.

The method I’ve landed on and keep coming back to is called CRIT. It’s widely available online and not proprietary to anyone, but it works for me because it mirrors how I think about framing problems. Here’s how I use it:

Context: Give the LLM the situational context you have. Tell it everything you know. Add transcripts of meetings, screenshots of text threads, project documents. The more context you provide, the better the output.

Role: What perspective do you want the LLM to take as it answers? A project executive reviewing a schedule? A client hearing your proposal for the first time? A superintendent evaluating constructability? Defining the role sharpens the response dramatically.

Interview: Have the LLM ask you a handful of questions, one at a time, to better understand your ask before it responds. This is the step most people skip, and it’s the one that makes the biggest difference. It forces the LLM to clarify your thinking before it starts generating.

Task: What you actually want to achieve. Be specific. “Help me write a recovery schedule narrative” is better than “help me with my schedule.”

Use AI as a thought partner.

This is where it gets powerful and where most people haven’t gone yet. Thinking through how to navigate onboarding a new employee into a difficult project environment? Trying to figure out the right approach for a tough owner conversation? Working through whether to pursue a project that’s outside your typical profile?

Throw your LLM into voice mode and talk it through. Give it the context of the challenges you’re facing, define what success looks like, and let it help you pressure-test your thinking. Use it to navigate opportunities, challenge your assumptions, and refine your approach before you act. It’s not replacing your judgment. It’s stress-testing it.

What’s on the Line

Your professional relevance.

That may sound melodramatic, but I stand by it. I believe there is a small window of time when the power lies with the individual to leverage AI before its value is captured by corporate.

Here’s why. Right now, AI tools are accessible, affordable, and largely ungoverned at the individual level. You can build your own workflows, accumulate your own context, and create genuine leverage in how you operate professionally. But that window is closing. Enterprise AI strategies are rolling out. IT governance frameworks are tightening. We’re already seeing companies formalize approved tool lists, restrict personal AI usage on company networks, and consolidate data into systems the individual doesn’t control. That’s not criticism, it’s smart business. But it means the room for individual agency inside corporate AI ecosystems is going to narrow.

AI is following the same arc, but faster and with higher stakes. Because unlike a project management platform, the value of AI compounds with context. The professional who has spent a year building memory, refining prompts, and integrating AI into their decision-making process isn’t just more efficient. They think differently. They see patterns faster. They operate at a level that’s genuinely hard to replicate.

The rational response isn’t to wait for your company to hand you an approved workflow. It’s to build your own stack now, while the tools are open and the guardrails haven’t been written yet. Capture your knowledge. Own your context. Create the leverage that compounds over time and travels with you regardless of where your career goes.

The risk isn’t in using AI. It’s in waiting until someone else decides how you’re allowed to use it.