Denis Rafailov

building business software for last 14 years. now I help them automate workflows with at least 3x ROI

[Denis Rafailov @denis_rafailov]

Remember, your business's growth often mirrors your focus.

The only two things you must do are

Find more Clients: Let them know what you do.
Deliver Better Quality: Listen, improve, measure, repeat.

Ideally every single member of your staff is either doing A or B.

[Denis Rafailov @denis_rafailov]

@_mctrinh If you train AI on AI generated data. On the 2nd or 3rd cycle the AI is complete scrap.
I will leave it to you to decide where the software industry is going

[Denis Rafailov @denis_rafailov]

@robj3d3 Don't rush it man. There is no point of burning tokens for the sake of having them. Using them would only matter if you would take the time to review what they produce.

[Denis Rafailov @denis_rafailov]

@Saanvi_dhillon git add .env

[Denis Rafailov @denis_rafailov]

LGTM! Absolutely amazing representation of tech companies now.
I hate how much I love it

https://t.co/1C29BsIeWf

[Denis Rafailov @denis_rafailov]

@TeslaFamBen You deserve it Ben! Congratulations 🍾

[Denis Rafailov @denis_rafailov]

@Samaytwt You really believe it or you are trying to push some buttons here 😅

[Denis Rafailov @denis_rafailov]

@TechOperator I go with APR (Artificial Pattern Recognition) If I have to add S I would go (Super Artificial Pattern Recognition)

[Denis Rafailov @denis_rafailov]

@TheLawrenceMac sounds painfully familiar :D Happy to connect. I can talk business or triathlon training all day long!

[Denis Rafailov @denis_rafailov]

@TeslaBoomerMama xD well yeah, you must follow him to unlock a follow back 😅

[Denis Rafailov @denis_rafailov]

@TeslaBoomerMama I challenge you to find such of Dario 🫣

[Denis Rafailov @denis_rafailov]

@JeenaKrish91074 You don’t return. Using the curtly brackets requires you to define a return statement. Otherwise if can also work if you remove them. It will be then interpreted as a statement

[Denis Rafailov @denis_rafailov]

I wouldn’t want to be the guy checking the code of a vibe coder. All vibe coding without fundamentals is experimenting. Once they screw the base you can’t fix it. It’s like painting a picture and going to professional to ask where to do the final touches. They would start with telling you why the foundation is wrong and how to achieve your concept but starting from a white sheet

[Denis Rafailov @denis_rafailov]

@jacobsharpuk @itsalicesoul Thank you Jacob! Happy to see there are still people with their minds out there. It’s crazy to see the other comments 😢

[Denis Rafailov @denis_rafailov]

@aslambeg84 * a to equal b (not 5)
Typo

[Denis Rafailov @denis_rafailov]

@aslambeg84 On line 5 you are assigning a to equal 5 which then would make that if statement to pass since the operation of assigning b to a passes. Which then would result to printing equal in any scenario

[Denis Rafailov @denis_rafailov]

I have been calling AI not be called AI but APR. Now I have to change it to SPR.

AI is great for helping us but it's misleading to call it intelligent. It knows and follows Patterns. It's being trained with Reinforcement Learning to implement Pattern Recognition.

That is why I call it APR and now I will SPR.

[Denis Rafailov @denis_rafailov]

Every business has TWO main ways of improving when integrating AI.

A: automate repeatable work
B: do something NEW

Both are important but one of these two you already do, and you know well.

In A: You have data to measure and understand what is your ROI. It helps you build a business case instead of relying on hope for it to succeed.

In B: You run an experiment. Something NEW always sounds good and shiny. Especially from a distance. But you have to understand you are running an experiment and you have to treat it as such. You have to build only the core feature, deploy quickly and iterate often. Don't spend a long time building when doing something NEW! This is the biggest mistake for any builder.

Overall one of these decisions has to be made only after you know your data. But don't just assume you know it all. MEASURE! Especially before making a change you would benefit way more from doing a two week Time-Study to show you the process in detail. Then use that data. Put it against the same process after you automated it.

This sounds like an ancillary step to the automation for many but this is the bridge that makes it bullet prove and more importantly shows you if it really works.

[Denis Rafailov @denis_rafailov]

Your best people are expensive for a reason!

They have the knowledge and judgment in the crucial moments.

That is why you pay them well.

But you are also occupying the same skills for low skill work.

And the better your employees are the more that waisted hour costs you.

So, when you depend on them for admin you are not only waisting money but also draining the focus you hired them for.

It's your responsibility to put systems in place that will not drag your team down but help them in their day-to-day operations.

Watch the full version here: https://t.co/WJ979DtTzD

[Denis Rafailov @denis_rafailov]

@SaraDiscovers Bigger membership plan