Hari Prasad Renganathan

Take-Two Interactive Software, Inc.

Data Scientist

• Contributed to 4 major Enterprise Data Science/ Analytics projects in the Sales team
• Collaborated with ML Engineers and Data analysts to deploy price elasticity outputs from AWS S3 buckets into production as an interactive dashboard using Dash
• Optimized sales-forecasting model by introducing ARIMA and improved performance by 33%
• Mined and analyzed data of 72 different companies to perform competitive analysis using hypothesis testing leveraging Snowflake, SQL and Tableau
• Performed market research of company’s 12 subsidiaries and presented the patterns to Marketing department

Data Science Intern

MyRealProduct

MyRealProduct is a 4-week cohort where aspiring Data and AI professionals build real, usable products instead of just toy projects, solving real-world problems across domains like healthcare, food, and productivity.

Mentor | Instructor

NYC Women in Machine Learning & Data Science 

Guest Speaker

The College of Westchester

ElevenLabs

Judged Voice AI Hackathon sponsored by ElevenLabs & conducted by Quru AI

Technical Judge

Globee Awards

Industry Judge

Google

I was invited as a speaker at the AICamp event, hosted at Microsoft NYC and sponsored by Google.

The energy in the room was electric! I dove deep into the world of AI content creation and explored practical ways to make AI writing sound more human. The engagement and questions from the audience were incredible - clearly this is a pain point many of us share.

Automate Apply

Delegate Applying For Jobs! Hire a Personal Assistant who can apply jobs on your behalf.

Angel Investor

BrowserStack

I was invited to give a speech at the BrowserStack event on "ML Model Testing & Validation."

When you think about model testing, what comes to mind?
Train-test split? 
Accuracy? 
Confusion matrix?

Yes! But model testing is more than just that.
It's about prioritizing business requirements and ensuring the model aligns with them.

I spoke about the importance of incorporating business context into model testing.

JotterWolf

Struggling to track your job applications?

JotterWolf addresses the common challenges associated with job applications and interview preparation. It offers the following benefits:
- Automated job application tracking
- Interview preparation support through ChatGPT extension
- Goal-setting functionality for maintaining job search momentum

Co-Founder

Associate Data Scientist

Bye bye America 🇺🇸

4.5 years ago, I landed in the US with 3 suitcases and a stupidly big dream.

Today, I’m leaving with… the same 3 suitcases.

But everything else inside me changed.

This country gave me
• My Master’s degree
• My first job
• My first startup
• My community
• My confidence

And still… I’m leaving.

Not because I want to.
But because of a lottery.

When my H1B didn’t get picked for the 3rd time, it broke me in ways I can’t explain.

You do everything right.
You work hard.
You build.
You contribute.
You pay taxes.
You stay out of trouble.

And a random draw decides your future.

Millions of people come to the US with a dream.
Some get lucky.
Some don’t.

I was one of the unlucky ones.

But here’s the twist life didn’t expect:

Endings create monsters. And I became one.

If the door doesn’t open, you don’t wait.
You build a new one.

So… new chapter.
New country.
New fight.

Hello London 🇬🇧
Let’s see what you’ve got.

America will always have a special place in my story.

But the story isn’t ending.
It’s leveling up.

Everyone saw my “Bye bye America 🇺🇸” post.
Some of you even DM’d saying you cried reading it.

But there’s one part of that story I didn’t talk about.

The months before that email.
The months where every international student lives in a strange limbo
not here, not there, not anywhere.

You work hard.
You build things.
You try to stay hopeful.

But deep down, every plan you make has an asterisk:
“Depends on visa.”

That mental load is something most people never see.

The confusion.
The late-night Googling.
The fear of choosing the wrong next step.
The feeling that your entire future is a “maybe.”

Recently, I’ve been doing something I never did before:
I started learning about my options properly.

Not through forums.
Not through panic.
But through real guidance.

And one team I’ve been speaking with is Manifest Law
I appreciate how they break things down without jargon. When you’re already drowning in uncertainty, simplicity matters.

If you’re on O-1 / EB-1 / NIW paths or figuring out what your next chapter could look like, you might find that helpful too. Or you can book a free consultation with them here: https://lnkd.in/gBu_sJcm

Because here’s what I’ve learned:
Your dream shouldn’t collapse just because the process is confusing.

Whenever life pushes you out of one door…
you’re allowed to walk confidently into the next one.

Immigrants don’t quit.
We adapt.
We rebuild.
We rise.
#ManifestLawPartner

We just crossed 45,000 people.

That still feels unreal 
So I wanted to quickly re-introduce myself and thank you.

Who am I?
✔️ I’m a Columbia University grad
✔️ A Data Scientist who transitioned into a Senior AI Engineer
✔️ I’m on a mission to create 365 AI Engineers in the next 365 days
✔️ I help professionals and businesses simplify their workflows using AI
✔️ And I try my best to motivate international students, because I’ve lived that journey myself

And who are we as a community?
✔️ We’re builders
✔️ We’re doers, not passive learners
✔️ We show up, experiment, fail fast, and improve
✔️ And somehow, we’ve grown into one of the fastest-growing AI communities out there

I just want to say thank you.

Thank you for being here.
Thank you for reading, sharing, supporting, and trusting me with your time.
This community is what it is because of you 
Every message, every comment, every person who decided to stay.

I’m genuinely grateful.
Let’s keep learning and building together. 🙏

"Here are the 4 things I did to get into Google 👇"

.....the classic LinkedIn post

Everyone wants a template for life.
A step-by-step guide.
A “do what everyone else is doing” safety manual.

But here’s the uncomfortable truth nobody says out loud:
The fastest way to lose in life is to copy everyone else.

I’ve never fit into templates.
And honestly… that’s the only reason I’m here today.

I come from Mechanical Engineering.
I work in AI now.
Already “off-template,” right?

People think I got into Columbia easily.
Nope.
I applied to 16 universities. Got rejected by many.
I even cold-emailed all of them asking for fee waivers.
12 said yes.
Applied for almost free.

When everyone was “just submit more applications” for internships,
I stopped applying.
Took a break.
Then I did something nobody around me did:
I texted founders directly on LinkedIn.
No fancy pitch.
No resume paragraphs.
Just: “Hey, I’m a Data Science grad. I want to help.”
Within a week:
10+ interviews.

People said,
“After graduation, you get a full-time job. Chill.”
But during OPT, I co-founded a company.
 
People said,
“If you don’t get H1B, it’s over. Pack your bags. Go home.”
I didn’t.
I got a UK job.
Asked them to sponsor me.
And now I’m moving to the UK with a Skilled Worker Visa.

You see the pattern?

Every big win in my life came from doing the thing people don’t do.

Not the safe thing.
Not the common thing.
Not the template thing.

So if you’re waiting for the world to approve your path…
you’ll be waiting your whole life.

Life isn’t a template.
Your career isn’t a template.
Your success isn’t a template.

And the irony? 
You probably came here expecting a “success template” from me.

The only template worth following is the one you create yourself.

So if you're an international student…
if you’re stressed about AI replacing your job…
if you're switching domains…
if you’re tired, rejected, confused, hopeful…

Please remember this:
It’s okay to be different.
It’s okay to take a break.
It’s okay to try the “weird” thing.
It’s okay to choose a path nobody around you is choosing.

Because sometimes the path nobody is walking
is the one meant for you.
And it might just change your whole life. ❤️

Everyone is busy yelling “AI will take your job.”
No one talks about the system that actually does it.
Immigration 🇺🇸

AI didn’t push me out.
AI didn’t limit my opportunities.
AI didn’t end my American Dream.

But a lottery did.

I got rejected by the H1B lottery 3 years in a row.

Not because I wasn’t qualified.
Not because I didn’t work hard.
Not because I didn’t add value.

Just because a random number generator said “Not selected.”

Imagine working your way into an Ivy League,
building AI systems,
joining a top gaming company,
contributing to real products,
helping hundreds of students…

And then one morning, an email decides your future.

No interview.
No discussion.
No appeal.
Just “Unfortunately, you weren’t selected.”

AI didn’t replace me.
Immigration did.

I’ll be honest, I felt stuck.
Frustrated.
Disappointed.
Embarrassed even…
like I wasn’t good enough for the country I gave everything to.

But here’s the part people don’t see:
Life isn’t a straight line.
It’s a series of brutal plot twists.

The same year I got rejected again,
I also got approved for my UK Skilled Worker Visa 🇬🇧
Suddenly, doors opened.
Opportunities expanded.

My career didn’t slow down
it accelerated.

Sometimes the door that shuts isn’t the end.
It’s redirecting you to a bigger room.

I know many international students and immigrants feel the same fear I once felt:
“What if I don’t get picked?”
“What if everything falls apart?”
“What if my dreams depend on a lottery?”
I’ve been there.

But trust me
Your story doesn’t end with a rejection email.
Life isn’t linear.

It’s a comeback story waiting to happen.
And you’re still in Act One.

You have 31 days until 2025 ends.
What if you could become an AI Engineer in the next 31 days?

“Hari, that’s bs.”
I know what you're thinking.

Save this post. You’ll want to come back to it.

Here’s a 31-day plan to learn AI
(Pre-req: Python)

Day 1 – LLM Basics
Day 2 – Prompts 101
Day 3 – Prompt Patterns
Day 4 – ChatGPT API
Day 5 – Build First App
Day 6 – UI + Backend
Day 7 – Deploy Simple App

Day 8 – Embeddings
Day 9 – Vector DBs
Day 10 – Chunking
Day 11 – RAG Basics
Day 12 – RAG App
Day 13 – Improve Retrieval
Day 14 – Memory + History

Day 15 – Tools & Actions
Day 16 – Function Calling
Day 17 – Multi-Step Flows
Day 18 – Agents Overview
Day 19 – Build an Agent
Day 20 – Add Tools
Day 21 – Decision Logic

Day 22 – Eval Basics
Day 23 – Quality Metrics
Day 24 – Guardrails
Day 25 – Monitoring
Day 26 – Logging
Day 27 – Fix Failures

Day 28 – Real Project
Day 29 – Polish + UX
Day 30 – Ship It
Day 31 – Portfolio Ready

If you follow this, you’ll know more than 90 percent of people “learning AI.”

If you want to do this along with me, I'm going to help 10 professionals to break into AI. 
Comment or DM "Workshop" 👇 I'll send you more details.

Everyone is talking about AI.
But almost nobody is actually building anything.

That’s why I’m doing something I’ve never done before.

I’m hosting The AI Innovators Summit 2025 on December 21.
Not a panel of “thought leaders.”

Not another event full of buzzwords and fluffy hype.
This is a room full of builders.

People who ship.
People who experiment.
People who actually do the work.

Here’s what we’re breaking down:
• How to become an AI Engineer in a practical, non-chaotic way
• How does Data folks use AI 
• What the top 1 percent actually do differently
• The exact skills that matter for 2026 and beyond
• A live Q&A where nothing is off the table

And the speaker lineup is crazy
builders & engineers from Google, Amazon Web Services (AWS), NVIDIA, Amazon, Walmart, and more.

Your favorite - Sohan, Venkata Naga, Vishakha, Karun & Irena (Thanks for your time & support. I'm sure this will help a lot of AI Enthusiasts).

Why am I hosting this?
Because I’m tired of seeing smart people feel “behind” when the truth is…
You’re one good framework, one good project, one good push away from changing your entire career.

📅 December 21, 2025
⏰ 11 AM – 12:30 PM EST
Free Virtual Event
Register here: https://lnkd.in/eETZPDmy

Spots are limited.
If you want 2026 to be the year you build, not just scroll…

I’ll see you there.
Let’s make something real.

Nikhil Kamath × Elon Musk - the podcast that shook the internet.
No, I’m NOT here to share “my learnings.”
The other 785 Linkedin Influencers shared it already (iykyk 👀)

I’m here to tell you the uncomfortable truth nobody is saying:

AI won’t replace everyone equally.
It replaces people in layers.
And you get to choose your layer.

-------- Layer 1 --------
People who don’t use AI at all

“AI is overrated.”
Still doing work the 2019 way.
They go first.
-------------------------

-------- Layer 2 --------
People who use AI casually

A few prompts.
A rewritten email.
A sparkly paragraph.
They go second.
-------------------------

-------- Layer 3 --------
People who use AI and fills in the gaps

They know when AI hallucinates.
They know where it fails.
They patch the weaknesses.
They survive longer.
-------------------------

-------- Layer 4 --------
People who help AI fill the gap

The builders.
The workflow designers.
The ones giving AI hands, legs, and tools.
They go last.
-------------------------

The formula is simple:
The farther you stand from the bottom, the harder it is to replace you.

Don’t use AI → replaced fast
Use AI → replaced next
Understand AI’s gaps → survive longer
Help AI fix its gaps → replaced last

Everyone says:
“People who use AI will replace people who don’t.”

Wrong.
That’s only layer 2.

The real winners aren’t the users
They’re the builders.

Not a big fan of Elon.
But this time I think,
Maybe Elon is right and one day we’ll all chill while robots work.

But until then?
Climb the damn layers.

Stop consuming. Start building.
Your future self will thank you.

Everyone is suddenly an “AI expert”… except you.

At least that’s how it feels, right?

Last week someone told me:
“Hari, I have no idea about AI. Everyone is learning it. I feel outdated.”

Let me tell you a secret nobody admits online:
Most people are pretending.
They talk in buzzwords.
They flex fancy models.

But if you ask them to build a simple working AI system
They freeze.

Because real AI engineering is not complicated.
There are only 4 levels. 
And it takes 1 week each if you stay consistent.

1️⃣ Level 1 -- Build a simple ChatGPT wrapper

Not a YouTube binge.
Not a Udemy marathon.
Just build 1 tiny AI app that solves 1 tiny problem.
That alone puts you ahead of 90 percent who “study” without building.

2️⃣ Level 2 -- Add RAG (Retrieval Augmented Generation)

Give your AI context.
Let it read your documents.
Give it memory.
Now it becomes useful.

3️⃣ Level 3 -- Agentic AI

Your AI can talk.
Good.
Now make it do things 
Make decisions, use tools, trigger workflows.
This is where you give the AI a body, not just a brain.

4️⃣ Level 4 -- Evaluate & Monitor

Most people stop here and celebrate.
Real AI engineers measure.
They monitor.
They improve the output quality.
This is the difference between a toy project and a product.

And that’s it.
No PhD.
No 1-year bootcamp.
No 100-hour playlist.
Just 4 levels. 4 weeks. 

Consistency > intelligence.

I’ve spoken to so many “AI engineers” over the last year.
You’d be shocked how few know even Level 2.

If you master these levels, you’re already in the Top 1%.
Not because you’re special.
But because you’re actually building, not pretending.

And if you feel behind, outdated, or overwhelmed…
You’re not.
You’re just one decision away from your comeback story.

I’m on a mission to create 365 AI Engineers in the next 365 days.
If you want to be one of them, follow my newsletter - https://lnkd.in/eveg5apm

Let’s build.
Let’s grow.
Let’s not pretend.

If you want to survive the AI wave in 2026, learn this simple rule:
AI Engineers don’t learn more tools. They learn fewer tools deeper.

People ask me every day:
“Hari… what should I learn to actually become an AI Engineer?
There are 500 tools. Where do I start?”

Here’s the truth nobody says:
The winners don’t chase tools.
They master foundations that make every tool easy later.

Here are the 5 skills you should master in the next 90 days if you want to stay relevant:

1️⃣ RAG Thinking → not just RAG building

Most people just follow tutorials.
Real AI Engineers learn how to:
 • chunk intelligently
 • pick the right embedding model
 • design retrieval strategies
 • run evals that actually improve quality
If you know this, you’re already ahead of 90 percent.

2️⃣ Prompt Trees instead of Prompt Dumps

One giant prompt is dead.
 The future is:
 • structured prompts
 • iterative reasoning
 • memory-aware prompt chains
 • fallback pathways
This is how you build reliable systems.

3️⃣ Workflow Orchestration
LLM + tools + condition checks.
Whether you use LangGraph, n8n, or your own Python scripts
You must know how to chain tasks like a real pipeline.
This is where simple builders become engineers.

4️⃣ System Design for AI Apps

You don’t need FAANG-level system design.
 But you do need:
 • data flow thinking
 • latency awareness
 • caching strategies
 • how to scale a vector DB
 • how to choose between serverless and containers
This alone will separate you from all tutorial-followers.

5️⃣ Shipping Small, Fast, and Real

Your portfolio is your new resume.
Not certificates.
Not badges.
Not "I’m learning AI."
Build:
 • one agent
 • one RAG app
 • one automation
 • one workflow

Just 4 projects. That’s enough.

If you master these 5, you will not be replaced by AI.
You will lead the people who are.

If you need help to break into AI, comment "WORKSHOP" 👇 I'll send you more details on how you can build along with me.

I’ve mentored 700+ people into AI engineering.
And almost all of them get 1 thing wrong.

They think becoming an AI Engineer requires:
• “research-level math”
• “10 years of experience”
• “learning 50 tools before starting”
• “understanding every model paper”
• “waiting until they feel ready”

Truth is… none of that matters.
NONE.

The only question that decides who grows 
and who stays stuck is brutally simple:

Did you build something this week?

Not perfect.
Not scalable.
Not YC-ready.
Not “beautifully engineered.”
Just real.

A tiny demo.
A scrappy prototype.
A broken agent that loops forever.
A RAG app that barely returns answers.
A workflow that only works on your machine.

Because the moment you stop being a consumer
(watching tutorials, saving bookmarks, taking notes)
and start being a creator
(shipping, testing, breaking, fixing)

Your career compounds like crazy.

I’ve seen it happen 100s of times inside my workshop MyRealProduct.
The people who win aren’t the smartest.
They’re the ones who touched something real, today.

If you’re reading this, here’s your sign:
Build one small AI app this week.

Your future self will thank you a thousand times over.
No fancy tech stack needed.

Just curiosity, momentum, and the courage to start messy.
Let’s build.

If you need help to build along with a community, comment or DM "WORKSHOP" 👇 I'll send you the details to join 
(It will be the best investment of your life. I promise!)

Albert Einstein said this long before:
• layoffs
• visa lotteries
• AI replacing jobs
• startups failing fast

But this quote hits harder in 2025.

Because stopping is the real risk.
Not failure.
Not rejection.
Not switching countries.
Not pivoting careers.

Stopping.

I’ve seen smart people pause because:
“I’m not ready yet.”
“The market is bad.”
“Let me wait for clarity.”

Clarity doesn’t come from waiting.
It comes from motion.

You move first.
Then things align.

Every big shift in my life came while moving:
 • learning while working
 • building while uncertain
 • posting while doubting
 • shipping before feeling ready

Momentum beats perfection. Every single time.

If you’re stuck right now, don’t overthink the destination.
Just pedal.

$3 million dollars....
Yes, you are reading it right 👀

We just unlocked $3 million dollars worth of perks for AI builders.
Yes, really.

Over the past few weeks, 
people inside MyRealProduct kept asking me the same thing:
“Do we get any credits or discounts for the tools we use to build our products?”

So… we made it happen.
Today, we’re launching MyRealProduct Perks 🎁

A curated set of discounts + cloud credits from tools you already use to build real AI products 
AWS, GCP, Azure, Notion, Perplexity, Make, and many more.

When we did the math, the total value blew our mind: 
over 3 million dollars in potential savings 🤯

This is our small Thanksgiving gift to the fastest-growing AI builder community 

A way of saying thank you for trusting us, building with us, and pushing real products into the world.

If you want check what perks we have :
https://lnkd.in/esPBVvZK

We’re not a “course.”
We’re a hands-on, product-first community for people who want to build real AI products, not just watch tutorials.

And if this is what we give away for free… imagine what we build inside
Happy Thanksgiving ❤️

Everyone says they’re “waiting for the right time” to start learning AI.

But here’s the truth nobody likes to admit:
Most people aren’t waiting.
They’re hesitating.

Because starting something new is scary.
Building something real is even scarier.

But you know what’s scarier than both?
Being in the exact same place next year.

Tomorrow is the last day to join MyRealProduct Winter 2025 Cohort.
Not saying you should join.

Just saying: whatever you choose, choose movement over stagnation.

Your future self will thank you for even the smallest step.
DM me or comment 👇 if you need the link.

taplio

Hari Prasad Renganathan's Linkedin Analytics