Founders are drowning in a sea of new AI tools promising to revolutionise their company’s approach to everything from product development and marketing to customer service. To benefit from AI though, entrepreneurs are going to need to quickly distinguish between genuinely transformative solutions and shiny (and fun) distractions.
Most founders don't need an army of AI assistants. As one founder told me: “I don’t want a tool that makes my customer service sound robotic or automated. Remaining customer centric is how we’ll win — not shaving off one team member in favour of an AI agent.”
So, what are founders actually finding useful? Here’s what three founders I spoke to are using.
Kaarel Kotkas, founder and CEO of identity verification startup Veriff
In the past, tools like Stack Overflow and Google were essential for founders to use to quickly build projects and find in-depth information. But you needed context to be able to do that — you had to become an expert in your area of interest to be able to sift through the noise of all the unorganised results.
Now, tools like ChatGPT and Claude have streamlined building and information processes — and you don’t need to be an expert to explore an area of interest. I can create prototypes, test hypotheses and refine our ideas quickly and with small teams. This is game changing. I personally use them as strategic thinking and validation tools. By sharing ideas with ChatGPT, discussing their viability and then going through iterations on the idea based on the feedback, my thinking evolves. This also applies to building prototypes.
For example, I can ask for a Python (coding) script that can be created and implemented into my projects right away. I can also ask what’s missing or what I haven’t thought of in my code. I programme and test things out daily with ChatGPT and it has sped up progress a lot. Of course critical thinking is the key here, but it creates quicker progress and enables me to prioritise based on evidence.
Now is the best time ever to be a founder but we need to be teaching these new skills to the founders and workers of the future. It’s why I’ve been involved in bringing AI tools into all schools for free in Estonia to ensure high school students learn AI skills — we need more personalised learning for students so that they can choose what areas of interest they want to go deep on and build things quickly, instead of just having a general education.
Nafisa Bakkar, founder and CEO of media platform and agency Amaliah
I spend a lot of my time in partnerships and sales meetings where rapport and relationship building are key. But taking lots of notes and then documenting them in a way that’s usable for the entire team is also essential in these conversations. AI tools can help you stay present, really listen and participate in the conversation instead of having your head down the entire time taking notes.
I use Read AI as my meeting copilot and it has given me so much of my time back. There are, of course, competitors like Granola and OtterAI. But I’m a big fan of Read AI. It takes notes for me, summarises conversations — for long chats, this is invaluable — and creates action points. It’s so useful to be able to refer back to what the client said their needs were and what we said we would provide. I can also send my team members a link to the summarised AI notes with mentions of where they need to take action. I should note it's not always perfectly accurate, so it's worth checking the summary after each call.
Tim Chong, founder and CEO of fintech Yonder
AI can’t fix all of your problems. The core fundamental problems like customer discovery, validation and figuring out whether you can build something a customer loves — especially at the early stages of building a startup — do not change in an AI-driven world. Those problems remain the same. What AI can do, though, is help you iterate through these cycles much quicker.
Founders should be using a tool like ChatGPT, Gemini, Claude, etc. for day to day workflows. For example, for customer research, AI tools can help you record and synthesise customer interview notes. It can also help pull out important and common responses to surveys.
For our engineers, AI is a useful co-pilot. They use an AI agent to explain their code line by line to help them debug — this is known as the Rubber Duck method, we’ve just upgraded it from an inanimate object to AI.
AI agents also help our non-coders write code, like SQL, to analyse their data without a programmer’s support. I also believe you can get value from vertical specific solutions.For example, if you need a quick landing page for testing, then Loveable is perfect. And the key to all use cases with AI agents is good prompting.
In the future, I would like to see a variation of deep research AI that can dig through all of your analytics data and pull out insights automatically. Like ChatGPT’s Deep Research, but you own your own analytics data.
