Justin Strharsky

Built 3 businesses to 8-figs revenue with community power. Now building a global community of AI creators and sharing what I learn.

[Justin Strharsky @JustinStrharsky]

I’ve helped more than 30 $5-15BN companies to achieve $MMs in efficiency, sustainability, and innovation gains, and built 3 businesses in the process.

But until 2015 I struggled to win any clients.

Then I discovered...

[Justin Strharsky @JustinStrharsky]

Could you accelerate your leadership career with a team of specialist advisors?
In this short clip, Kristina Kaganer shares how she created an always-on “LLM advisory board.”

[Justin Strharsky @JustinStrharsky]

AI leaders: want to make stronger project pitches and run fewer dead-end pilots?
In this 60-second clip, Kristina Kaganer reveals her go-to “sparring partner”: a personal board of LLMs. Before she puts forward an idea, she asks ChatGPT, Gemini, and Claude to rip it apart.

[Justin Strharsky @JustinStrharsky]

Constantly hearing about the AI talent shortage?
Maybe we are missing something right under our noses?

Sam Fletcher’s team uncovered 51% more internal data science skills.

Result: a clear map of previously hidden skills and internal mobility opportunities that most overlook.

[Justin Strharsky @JustinStrharsky]

Why did this Head of Data Science refuse to lead projects?
When Adrienne Leonard was Head of Data Science at a major aviation technology firm, she insisted that her team would never 'lead' a project.
Here's why:

[Justin Strharsky @JustinStrharsky]

Good AI leaders don’t just delegate; they prototype.
According to Peter Gostev, it's critical to work hands-on with the latest AI models.
How else can you know what's possible and which use cases are worth building?

[Justin Strharsky @JustinStrharsky]

What's the best operating model for AI? Should GenAI be in its own team forever?
Peter Gostev argues “probably not.”
Right now the tech is changing too fast, so a dedicated team is the only way to go.
But once the tech stabilizes, expect GenAI to collapse back into engineering.

[Justin Strharsky @JustinStrharsky]

Do you have to be tri-lingual to lead an AI team?
You've got to balance 3 things, says Merelda Wu: business goals, experimental science, and production‑grade engineering.
Great leaders speak all three languages and translate in real time.

[Justin Strharsky @JustinStrharsky]

What is a tree-shaped engineer?! 🌳
Merelda Wu dropped this idea in our conversation and I haven’t stopped thinking about it.
It’s one of the best ways I’ve heard to think about building great dynamics on AI teams.
Watch the clip, maybe you'll start growing a forest.

[Justin Strharsky @JustinStrharsky]

Humans are funny.
Preaching big vision. Sweating small stuff.
I talk a lot about abundance. It’s how we help companies solve the AI talent crisis in a novel way at Humyn-ai.
And yet, I catch myself agonizing over whether I really need one more $19/month SaaS tool.

[Justin Strharsky @JustinStrharsky]

@radekosmulski @kaggle Well done Radek! 💪

[Unearthed @Unearthed_S]

Want to join Australia's most influential mining event and represent your country? 🌏

We are on the lookout for the brightest minds and boldest ideas from around the world to take the mining industry to unprecedented heights.

Submit your tech here: https://t.co/s8hJTfD3HB

[Santiago @svpino]

90% of you will read this post, bookmark it, and forget about it immediately.

But more bookmarks won't help you become a better data scientist. To make progress, you must do something different.

Here is one idea:

@humyn_ai is a platform where you can work on real projects for real companies. You don't pay anything and get access to the following:

1. Real-world datasets to solve real-world projects.
2. You'll meet other people like you and work with them.
3. These projects are perfect for building your resumé.
4. You can win cash for your work if you do well.

Every day, many people ask me how they can get a job without experience. Other people are stuck solving the same problem for years without a way to diversify outside their box.

A platform like @humyn_ai is the solution to these problems. They connect people like you and me with companies with a dataset and a problem to solve. This is how you can try and learn from a different industry, how you build a curriculum, and gain experience.

This is your opportunity!

Go to their platform and sign up. They just released two new projects you can join today.

[humyn.ai @humyn_ai]

Curious about #largelanguagemodels and #GenerativeAI? Ever wanted to roll your own?

New opportunities coming soon! Make sure you sign up to hear about them first: https://t.co/4kS1gdyOuw

[Justin Strharsky @JustinStrharsky]

@volodarik Here's a choice line from ours: " Welcome to Humyn-ai, the website where humans pretend to be intelligent! Don’t be fooled by the name, there’s nothing humane about exploiting cheap labor from around the world to do your data science projects for you." Burn! 🔥

[Justin Strharsky @JustinStrharsky]

@volodarik Maybe it has too much content? 😉

[Justin Strharsky @JustinStrharsky]

@nathangloverAUS Nice!

[kirsten lum @kirsten_lum_]

Why do ML projects fail? It’s not usually about technology. It’s about people.

👩🏻👦🏽👨🏼‍🦲🧑🏾‍🦳👩🏿‍🦱👨🏼👱🏾👱🏽‍♂️👩🏼‍🦳👨🏿‍🦳👩🏼‍🦲🧑🏾‍🦲🧔🏻‍♂️

First, let’s talk about successful ML projects. Almost invariably, they bring together a group of people with different but overlapping expertise:

<1/13>

[Justin Strharsky @JustinStrharsky]

Anyone using GPTChat for data cleaning and transformations?

[Justin Strharsky @JustinStrharsky]

@MillyTamati Congrats Milly!

[Justin Strharsky @JustinStrharsky]

Are you selling to enterprises?

How many times are you asked about IP, freedom to operate, terms, team, certificates of insurance, bank letters, ultimate ownership, conflicts of interest statements...

Have you used a data room for enterprise deals before?