So, do you want to build a moat?
Most AI products today are extremely superficial. We can go deeper. 2026 is not about building minimum viable products, it is about building minimum lovable products.
95% of Startups out there are sticking out a few prompts together, and expect to automate 100% of a use case with minimal testing. When the sales team shows it to a prospect, if they have decent engineers they can probably build it themselves in a few weeks.
I’ve only been impressed by a few companies that are leveraging technology in the right way.
Example of a successful company: Lovable. The magic is not to throw some prompts and get an application. It simplifies so much more. Before Lovable I had to:
Host a website and configure it to see results. 30 minutes of boring admin work.
Download a code editor and connect it to an LLM. Input my credentials for it. 5 minutes more of annoying admin work.
I had to specify which files I want edited and use the terminal. Every change I need takes me like 1 minute and a few clicks.
Lovable made this very easy, even for non technical users. Heck, you can even talk to it now.
We need to make winning procurement contracts this easy.
There is a similar European startup with a lot of traction. They have 1 million users and only 4 employees. It’s called AiApply, and within 1 month you can apply to 500 jobs without having to do anything besides giving them your resume and something like $200.
So we have agents that can login to sites that they’ve never seen before and actually fill out complex application forms that require 2 factor authentication and take multiple questions.
As long as they have the right context.
My realization
I was doing a discovery call this quarter with a prospect. He mentions that they are currently paying 50K per year with a competitor. I showed him Tendermarket and he says that our application looks exactly like what they are currently using, except that we have “Market Insights” and he is curious about it. We exchanged a few emails, but the conversation didn’t go anywhere. The problem wasn’t big enough for him. He is not going to switch from a 2 year old locked contract for a “nice to have” feature.
He showed me a video from another founder's solution, they were building this AI agent that generates a Gantt Chart for the proposal. Very different from the other solutions, but he was excited about it. It made an impact on him.
My prospect still spends maybe 2 hours everyday using ChatGPT to match their tender requirements, he has to login, find proposals and keep track of all documents and deadlines, some tenders require a Gannt Chart. Think about his workflow. There are still a bunch of clicks he needs to do, multiple logins, pulling documents together, signing some and creating some more.
All these actions are perfect use cases for AI.
I’m wearing my Product hat, what he actually meant to say is not that he wants a Gannt Chart, he wants a solution that makes his life easier.
Our AI agent in action pulling tender requirements and questions live. I haven’t heard of any AI procurement company doing this yet, at least not in a way doesn’t require the user to upload the documents manually with RAG.
After testing a bunch of web browsing agents, and putting some dozen tools together, we came up with a solution that logins to the biggest European procurement platform, downloads all tender requirements and understands the questions. This takes our automation 15 minutes instead of 2 hours of a human playing around asking questions to the LLM, and uploading documents.
This is impressive in itself. Once we tweak this solution to automate the 40 hour process for creating a full bid, we will be able to impress every single procurement team.
The hardest part is done.
INTRODUCING OUR NEW BOARD MEMBER - JESÚS ALONSO GALLO
Jesús is a startup veteran that has invested in 82 startups, and created 4 companies with 3 successful exits. He sold restaurantes.com to the Michelin group.
I highly encourage you to subscribe to his newsletter. Yes, it is in Spanish so use Google Translate because it’s worth it. I could listen to him all day.
Jesús will be helping us with our Go To Market strategy and with a few introductions to his network where he thinks there can be synergies.
Jesús decided to dedicate the rest of his days to mentor entrepreneurs and writes everyday on his newsletter FIVE about Happiness, Investment, Entrepreneurship and Life.
VISUALIZING THE FUTURE
Summarizing my vision from Tendermarket.eu. If we can bring down the cost of making a bid from $5000 to $1000, we will become too big of an advantage for any company to ignore, while maintaining profit margins of +90%.
Onwards, Juan