From the Top: Learn the Instrument Before You Write the Song

Recently, I have been encouraged to see more team members at TEECOM engaging with AI tools, exploring how they might solve real problems in our workflows. That initiative is exactly the kind of energy a company at the leading edge needs. It reflects curiosity, ownership, and a willingness to lean into change.

But I want to offer a course correction on how we direct that energy, because the difference between productive and unproductive effort in this space comes down to one question: Are you learning the tool, or are you building around it? ‍

The Build-First Instinct

When engineers encounter a problem, our instinct is to engineer a solution. That instinct has served us well for decades. But in a rapidly shifting tool landscape, it can work against us.

Here is the pattern I am seeing: someone identifies a genuine pain point in their workflow, spends time scoping a custom solution, and builds something that addresses that specific problem. The work is often impressive. But the underlying workflow may soon change. The problem was real, but it was not as big or as permanent as it felt in the moment.

When the tools and workflows you utilize are changing quickly, a custom-built solution aimed at today's pain point can become obsolete before it delivers value. That does not mean the effort was wasted. It means the effort would have been better spent learning the tools that are reshaping the workflow itself.

At TEECOM, I am asking everyone to recalibrate this sequence. Learn the tool first. Understand what it can already do. Then decide whether you need to build something custom.

The Better First Move

The single highest-leverage thing you can do right now is not to build an AI-powered tool. It is to use AI tools to accelerate your existing work.

This is where TEECOM's investment in GitHub, markdown, and structured documentation is starting to pay real dividends. Every markdown file in our documentation repository is machine-readable. Every template, onboarding guide, process document, and standard we have published in GitHub is already in a format that AI tools can search, summarize, and reason about.

Utilizing existing AI tools, we can query our entire documentation repository using natural language. Our teams can ask it to find the relevant process for a task, summarize our onboarding process, and ask it what our documentation says about a particular problem. That is not a future capability. It is available today, and it works because we made the decision years ago to write in plain text and store our knowledge in version-controlled systems.

The reason we had our teams learn markdown and GitHub was about making our collective knowledge accessible to the tools that are now arriving.

What You Can Do

Before you spend time building a custom tool, spend time learning what the tools your team has access to can do. The gap between "I have a problem" and "I need to build something" should include a step most people are skipping: "I need to build my skill level using the tool across the spectrum of tasks I perform."‍

Here are two concrete steps you can take this week:

1. Before starting any new tool or automation project, write a one-paragraph problem statement and send it to your team. Get feedback on whether the problem is already being addressed by an existing initiative, or whether a simpler approach exists. This takes five minutes and could save you days.
2. Spend thirty minutes this week exploring one way you have not used an AI tool before. Not building with it. Just using it. Have it review a document. Use it to understand a piece of code or a spec section you did not write. The goal is fluency, not a finished product.

The Bigger Picture

AI tools are evolving faster than any technology shift I have seen in my career. The workflows we use today will look meaningfully different twelve months from now. That is not speculation. It is already happening within TEECOM.

In that environment, the most durable skill is not the ability to build a bespoke tool. It is the ability to quickly learn and apply new tools. The people who develop that habit now, who become fluent in prompting, querying, and collaborating with AI, will have a compounding advantage over the next several years.

Learn the instrument first. The songs will follow.

David Marks, CEO

‍