Many companies believe that having a massively complex hiring process is synonymous with quality.

Most people still don't realize that an LLM is just a text-completion model, lacking intelligence and any commitment to the truth, and because it's trained to give specific answers, it's a hotbed of bias! It pains me even more to see the term AGI

The biggest mistake I ever made was ignoring advice from more experienced ones

Think about how you act—are you efficient, or do you react without thinking?

since cursor and the Claude code was available, I never ran out of tokens, but after I saw how people using AI to code, I understood why, hahaha.

The use of the term "AI" these days bothers me—it's depressing.

What would happen if you minimized backend development, where data modeling, business logic, and everything else (like auth and all CRUD operations) were automatically generated? Not only could you build massive, complex systems in minutes, but you would also use 100 times fewer tokens. This is entirely possible today, ask me how if you're interested... though the only limitation is that it currently works only with the Node.js + Postgres stack.

I can finally say I’ve learned how to validate an idea that people are willing to pay for, that’s the skill founders most often lack.

Validating the business and setting up the operation is more difficult than introducing technology to your service.

I created a backend where all the logic lives in the database (postgres/sql); Node.js is only used to interface with the REST.

Just sharing something with you here: this month I closed a $300K contract to develop a casino product, with $200K going to the guys who closed the deal and $100K coming to me (gross), I’m not one to share this kind of information, but when I mentioned it to another “entrepreneur,” he said he thinks this split is absurd, given that I’ll be developing everything myself with my team. He just forgot that closing a deal like that is a hard as hell. Sometimes you don’t grow because you can’t recognize the value of others’ work. Another thing is, the contract could have been for a million... I only care if the pay that ends up in my pocket is attractive, not how much the other person is making.

The CEO must never be a bottleneck; they must be the enabler.

One day, people will laugh when looking back at the time they used ChatGPT with prompts like "act like a professional in [field]," thinking the LLM actually "knows" what it's doing instead of leveraging their own years of empirical knowledge to guide it toward what they already know is effective.

I worked for a USA bank that spent over $80k on AI traffic optimization tools, spearheaded by a growth lead (got fired) who never even looked at the metrics; according to Ahrefs, only 0.02% of the traffic came from LLM providers, while basically everything else came from Google. I argued that the best investment would be trying to improve page conversion, but they ignored me.

Am I the only one who thinks these AI visibility tools are crap? LLMs are non-deterministic by nature and rely on what currently exists on the internet (SEO); AEO from an organic standpoint depends on SEO.

If your SaaS isn't selling, share the link here and I'll evaluate it for free

I'd say spam.

One of the most common problems when marketing your SaaS is focusing on features; most of your customers won't fully grasp that, so you need to shift your approach to something simpler to understand. I like to explain the benefits that is, the problem the feature solves, rather than the feature itself. Imagine this scenario: you don't know anything about car mechanics. One salesperson tries to sell to you by citing engine specifications, while the other sells by saying things like "this car is faster".

One thing I realized and which was my turning point after hitting a lot of roadblocks, was to shift my focus to B2B. B2C is good, but your customer acquisition cost has to be low (among other issues); sometimes you just need software with 20 or more good-paying customers to give you financial security.