The Data Entrepreneurs

Founded a community of 850+ data & AI entrepreneurs. Hosting monthly workshops and discussions.

Founder

Shawhin Talebi Ventures LLC

Teaching AI on Maven. Shipping products. Advising 1-1k person businesses on AI.

AI Educator & Advisor

Towards Data Science

Writing articles on Data Science + AI. 25k Monthly Reads. 16k Followers.

Writer

YouTube

Making videos on Data Science + Entrepreneurship. 1M+ Views. 40k+ subs.

Content Creator

Toyota Financial Services Corporation

• Uncovered and corrected issues in production credit risk model that impacted over 70% of accounts and wrote model monitoring scripts to help avoid future failures
• Redeveloped loan originations model for independent dealers, which resulted in a 50% model performance improvement and provided $2.5 million in realized value to business partner

Data Scientist

The University of Texas at Dallas

• Published open-source methodology to discover optimal EEG bands which led to better characterization of the underlying power spectrum than the more commonly used band boundaries by a factor of 2
• Train over 100 machine learning models which estimated particulate ma]er (PM) concentrations based on a suite of over 300 biometric variables which accurately estimated PM concentrations with high fidelity (r2 = 0.91)
• Led 5 teams focused on deployment of a real-Ime Python-based biometrics application, unveiling immediate insights that were previously inaccessible

Research Assistant

Palomino Motors

• Analyzed marketing and sales reports to inform inventory acquisition, which resulted in a 50% decrease in average inventory age
• Evaluated costs of lead providers through analysis of lead data and close rates, which led to over $2500 in monthly savings

Business & IT Manager

•      Hired, trained, and led team of 8 SUAAB executive council members, that planned and executed 36 events with over 14,000 in total attendance 
•      Managed and analyzed programming budget of over $216,000 along with attendance reports to inform future campus events

Executive Chairman - Student Union and Activities Advisory Board

•       Orchestrated and implemented 2016 Spring Concert with over 1200 students, 45 volunteers, and 6 outside companies and organizations at the TI Plaza, that is now used as a model for all Spring Concerts
•       Analyzed aggregate student course schedules to predict optimal event times

Music Chairman - Student Union and Activities Advisory Board

Failing my way to success | Data Entrepreneur

Meet Shaw, an educator, data scientist, entrepreneur, researcher, drummer, and bread enthusiast driven by an insatiable curiosity to understand the world and create better systems.

With a Ph.D. in Physics from The University of Texas at Dallas, Shaw’s research focused on novel AI applications to fields spanning neuroscience, environmental science, and human performance.

As an educator, Shaw is driven by a passion for learning and a mission to make AI accessible to all. His YouTube channel and technical blog have taught millions and have gained an audience of over 50,000 learners.

Before you “vibe code” your next AI app… consider this.

With today’s powerful LLMs, almost all coding work can be offloaded to an AI.

While this helps developers move 5X faster, this speed comes at a cost.

Just like how taking a loan to launch a startup incurs financial debt, using an LLM to code your application incurs technical debt.

In other words, you trade speed today for work you must do later.

Technical debt (like financial debt) isn’t inherently bad for a business; in fact, it can be very beneficial when used appropriately and skillfully.

Here are the 2 key things I consider when deciding how to use LLMs to code.

1) My Experience

Vibe coding a binary classifier is much less risky for me than vibe coding a website with React.

That’s because I’ve trained dozens of classifiers before but never used React.

2) App Run Count

Vibe coding an ad hoc data dashboard has very little risk.

However, vibe coding a SaaS app that will be used thousands of times by users presents a high risk because each run is an opportunity for a bug to appear and wake you up at 3AM to fix it 😅


How do you think about LLM’s writing your code?

--
If you liked this post, repost it and follow me (Shaw Talebi) for more :)

Shaw Talebi

Today, I posted the 2nd video in my new series on AI agents!

It discusses a central feature of agentic systems: tool use

Tools transform LLMs from word predictors into systems that can interact with the real world and solve problems on our behalf.

In the video, I share a concrete example of how to make a simple YouTube video agent using OpenAI's new Agents SDK.

👉 Check it out: https://lnkd.in/g4uzP7rG

PSA for (aspiring) content creators:

"You can’t eat followers!"

As someone who uses content creation in my entrepreneurial efforts, my biggest mistake was giving too much weight to follower count.

This all stemmed from a misunderstanding of how content creators (actually) make money.

While this comes in many flavors, they all boil down to two things.

Way 1: Sell other people’s stuff

This is what most people think about when it comes to monetizing content.

It includes: Medium’s Partners Program, YouTube Ad Sense, brand partnerships, and affiliate links.

Way 2: Sell your stuff

Although Way 1 is great for someone who wants to become a “creator”, many people I talk to don’t necessarily want content creation to become their main job.

Common things you can sell as a technical creator are: paid 1:1 calls, freelance or consulting services, digital products (e.g. ebooks, spreadsheets, code), a course, and a SaaS.

It took me 4 years to realize this... Hopefully, this post helps you realize it faster 😅

--
If you liked this post, repost it and follow me (Shaw Talebi) for more :)

Dimitri Visnadi is a data scientist turned independent consultant. In this podcast, we’ll discuss the top questions new freelancers ask and how to launch your independent consulting business in 2025.

👉 Register here: https://lu.ma/zqg4x3p1

How to Become a Data Freelancer (in 2025) | TDE Podcast #18

Hamel Husain is an ML engineer with over 20 years of experience and an AI consultant at Parlance Labs: https://parlance-labs.com/

Rapidly Improve AI Products with Evals | TDE Podcast: Hamel Husain

7 AI use cases that have helped me avoid hiring people (as a solopreneur)…

The barrier to entry has never been lower for entrepreneurship, thanks to AI.

Here are a handful of ways I use it every week.

1) Coding assistant

Cursor has accelerated my development time by an order of magnitude.

What used to take me a week to code, I can now do in a day.

2) Going from 0 to 1 on a new topic

Entrepreneurs need to be generalists. This means knowing the basics on a wide range of topics and being able to learn new topics quickly.

ChatGPT is my go-to for this. For example, I recently needed to learn how to track traffic in my marketing funnel, which introduced me to various “conversion tracking” tools.

3) Writing landing page copy

As a data scientist, I knew nothing about copywriting (and still know little).

However, with ChatGPT's help, I recently improved the copy on various landing pages, which has already shown a significant boost in conversions!

4) Writing email templates

Another area in which I am entirely inept is sales outreach.

This is why I used ChatGPT to refine my warm outreach scripts when promoting my previous AI cohort.

5) Crafting idea customer profiles from reviews

A central part of entrepreneurship is understanding your ideal customer.

Rather than do this manually, I passed all my 5-star reviews to ChatGPT and refined the avatar for my AI course.

6) Transforming my podcasts into blog posts

I have a podcast where I talk to technical entrepreneurs about their journeys and key lessons learned.

When a particular episode does well, I transform it into a blog post using a tool I built called y2b.

7) Creating peer groups for my course

I started assigning peer groups for my AI cohorts to promote engagement and collaboration.

Since doing this manually would take me 15 min or so, I just upload people’s experience levels and preferences to o1 and have it form them for me in seconds.


What are your top AI use cases for business?

--
If you liked this post, repost it and follow me (Shaw Talebi) for more :)

Taplio

Shaw Talebi's Linkedin Analytics

's Best Posts (last 30 days)

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Shaw Talebi's Linkedin Analytics

's Best Posts .css-qqfgvy{font-size:var(--chakra-fontSizes-md);}(last 30 days)

Want to drive more opportunities from LinkedIn?

Shaw Talebi's Linkedin Analytics

's Best Posts (last 30 days)