Build your minimum viable AI toolkit

Mithun Das
4 min readJan 26, 2025

Do you feel like you are falling behind in the AI race? Do you feel like you must know all the AI tools popping up to stay in the race? Trust me, you are not alone.

In today’s rapidly evolving AI landscape, it seems like every other week there’s a new large language model (LLM) or AI tool promising to revolutionize how we work. New benchmarks showing how the new model outperforms every other model out there — a new state of the art model going to take all the market by storm. But trying to keep up with all of them can be too much.

While innovation is exciting, the sheer volume of new tools can overwhelm even the most tech-savvy professionals. If you are a developer — you must be thinking should you use Github Co-Pilot or Cursor AI or Windsurf or Ripple? You pick one and next week you hear a new release from the competitor that you must use and ready to make a switch. Well, you can do that but you may be losing time to learn a new tool all together and the funny thing is by the time you are starting to get productive on the new tool, your previous tool rolled out the same feature — to stay in the race you effectively decreasing your net productivity.

I personally feel — it’s simply overwhelming. That’s why I have built a strategy which I call “minimal viable toolkit”.

👎 Why Overwhelm is the Enemy of Productivity

Every new AI tool or LLM that enters the market comes with a learning curve. Even if a tool boasts incredible capabilities, adopting it often means redefining workflows, reconfiguring integrations, and spending time understanding its nuances. These frequent shifts can result in diminished productivity and lost time — time that could have been spent mastering a tool that already works well for you.

🛠️ Build Your Own Toolkit

Instead of jumping on every new AI tool, focus on building a toolkit that serves your specific needs. This isn’t about finding the most advanced tool on the market; it’s about finding tools that fit seamlessly into your workflows and solve real-world problems. Here’s how to do it:

  1. Evaluate Your Needs: Identify the areas in your workflow where you need the most help. Look for tools that specifically address these gaps rather than general-purpose solutions that may be overkill.
  2. Try and Test: Experiment with a few AI tools and LLMs to see which one integrates best with your current processes. Evaluate their performance not just on benchmarks but in real-world scenarios.
  3. Commit to Mastery: Once you’ve found a tool that works, commit to it. Stick with it long enough to master its features and understand how to leverage it fully. This consistency often yields better results than frequently switching tools.

⚖️ The Balance Between Sticking and Switching

Well, it comes with a trade-off and you need to understand that. While consistency is key, it’s also important to stay informed about the latest advancements in AI. The goal isn’t to stick to a tool forever but to make informed decisions about when to upgrade your toolkit. Here’s how to strike the right balance:

  • Stay Curious: Keep an eye on new tools and technologies. Subscribe to newsletters, attend webinars, and follow AI thought leaders to stay updated.
  • Be Strategic About Switching: Only consider a new tool if it offers clear, tangible benefits over your current setup. Evaluate whether those benefits outweigh the time and effort required to adopt it.
  • Continuous Learning: Dedicate time to upskilling and learning about emerging trends so you’re prepared to pivot when the time comes.

Live as if you were to die tomorrow; learn as if you were to live forever — Gandhi

🧰 My Choices for AI Tools

For coding assistance, I picked Codeium Windsurf over GitHub Copilot and Cursor AI because of its agentic workflow when I started. While other tools now offer or are planning to introduce agentic workflows, I’ve stuck with Windsurf. I’ve learned to make the most of it, and unless there is a big benefit to switching, I’ll continue using it.

For video editing, I chose FlexClip.com over Synthesia and InVideo because it offers the right balance of cost and features.

When it comes to search, I prefer Perplexity over Google Search for its conversational nature.

For research work NotebookLM is my go-to tool.

🧠 Final Thoughts

The minimal viable toolkit strategy is not about doing less; it’s about doing more with less. By focusing on a handful of tools that truly serve your needs, you can avoid the constant churn of adopting new technologies and instead achieve mastery and efficiency. At the same time, staying informed ensures that you’re ready to make changes when they truly matter.

In the end, the goal isn’t to chase every shiny new tool but to build a robust toolkit that helps you work smarter, not harder. Stick with what works, stay curious, and always be ready to evolve — but only when the time is right.

Let me know if you have build your own strategy and like to share with others.

--

--

Mithun Das
Mithun Das

Written by Mithun Das

Software Engineer | Designing & Building Softwares for 20 years

No responses yet