Z47 Moments

What do you do when you've built the technology first and the first customer who tries it tells you it's not working at all?

Prashant Warrier is the first PhD and the first scientist on Unstarted. A steel-town kid from Bhilai who went from IIT Delhi to a PhD in AI, he came back to India because of a stolen passport — and built Qure.ai, whose AI reads chest X-rays for TB and lung cancer across the world.

In this episode, Avnish and Prashant attempt to answer:

 1. How do you identify a unique problem is it about passion or market cap?

2. How do you know the technology is actually working?

 3. How can founders assess whether they're early, late, or correctly timed for a market?

 4. Will frontier AI kill vertical businesses or is there domain-specific stuff left to build?

 5. Why build a cutting-edge AI company from Mumbai?

This episode is in partnership with TEAM — Tech Entrepreneurs Association of Mumbai.

Chapters 

00:00 Introduction
1:41 Steel town to IIT to a PhD in AI
3:20 SAP, and the stolen passport that sent him home
6:32 Freedom over the US visa rat race
8:17 The 2012 breakthrough that started the company
11:43 A research lab with no business plan
13:33 "A solution looking for a problem"
16:52 The reality check & the real bottleneck
19:20 TB screening: 4 weeks to 30 seconds
21:43 Detecting lung cancer early with AI
24:23 Moats, ChatGPT & the future of AI in healthcare
32:04 Why Mumbai, and advice to founders

He built a ₹100 crore brand from scratch in two years. His company rated him 2 on 5. So the question stopped being am I good at this? and became am I even the right kind of person for it?

Shashank Mehta is the founder of The Whole Truth, one of the best consumer brands built in India in the last decade. We backed him early, off the strength of a blog whose readers became the company's first customers.

In this conversation:
 
1. What do you do when you build something from zero and still get rated below the person who grew an existing brand by 5%?
2. How do you tell the difference between being bad at a job and being the wrong category of human for it?
 3. Is it risk that stops you from leaving — or is it fear wearing risk's clothes?
 4. Why does writing your worst fear down on paper shrink it?
 5. Can a single founder build a company that outlasts the founders who have co-founders?

A conversation about self-knowledge as the real moat — and why the fear that freezes you is almost always smaller on paper than the life on the other side of it.

Chapters 

0:00 Welcome & introducing Shashank Mehta
3:00 Middle class Delhi, mom's tears and the pressure to study
8:00 HUL, the startup bug & discovering he's a creator
14:00 First leap — joining Fasos, failing and coming back
20:00 The blog, the rage and the protein bar made at home
27:00 Risk vs fear — writing down what actually scared him
33:00 The 2/5 rating that built a ₹100 crore brand
38:00 Co-founders — why chemistry beats skill every time
43:00 Brand first vs revenue first — the 20 year argument
47:00 Building culture on trust and inspiration not fear and greed

Nirja Bhatt looks intimidating on paper: Columbia, Harvard Business School, strategy consulting. But she calls herself "a duck in water” i.e you never see what’s moving beneath the surface. It took her two years of analysis paralysis, ten abandoned ideas, and learning to normalize embarrassment before she could finally start.

Nirja is the founder of Laani, a pre-launch personal care brand for women, built around the categories nobody talks about. In this episode, she and Avnish go back to the very beginning.

They get into: 

1. How do you start when you have ideas but no capital? 
2. Why validate a problem, not an idea? 
3. How do you make peace with betting your own savings on yourself? 
4. Should you raise money before you have a product? 
5. And what actually separates the founders who make it?
6. An honest conversation about the part of the founder journey nobody photographs.

Chapters

00:00 Intro: Meet Neerja Bhatt, founder of Laani
01:30 Growing up in Baroda, raised by grandparents
03:00 Rejected by every dream school: the USC detour
04:30 How a professor's letter got her into Columbia
07:30 Q1: How do I start with no capital and too many ideas?
08:30 Analysis paralysis: 10 ideas and the HBS trap
11:00 Moving back to India: the waste management chapter
16:30 Q2: Passion vs. market gap — how to pick your problem
19:00 Talking to 200 women: customer validation done right
21:00 Q3 (Gubbachi founders): Hero products vs. commodity staples
26:30 To raise or not to raise: funding pre-launch
29:00 Final advice: persistence and eating a lot of shit
30:00 Outro

Chakradhar Gade had the résumé everyone's supposed to want: engineering, CFA, a hedge fund, and felt like he'd lost himself inside it. So he walked away to sell milk.

This is the story of how a finance guy who once mapped "the IRR of a cow" learned that the spreadsheet was a lie, lost all his money proving it, and rebuilt Country Delight from first principles, one customer relationship at a time.
Avnish Bajaj and Chakradhar get into the questions most founders sit with alone: 

 1. What do you do when a successful career leaves you with "a huge loss of identity"? 
2. Should you bootstrap and survive, or raise aggressively and run? 
3. How long does it really take to learn a business you've never been in? 
4. How do you tell real customer love apart from people just being nice to you?
 5. When the model that "looked very beautiful in Excel" collapses, what replaces it?

A conversation about chasing meaning over status, why bootstrapping bought six years of depth, and why, in Chakri's words, "there's no downside" to starting.

Chapters 

0:00 Welcome & introducing Chakradhar Gade
1:35 Growing up in Guntur, Infosys & chasing meaning
4:43 From MBA to Wall Street — why he chose finance
7:33 The decision to quit New York and become a doodhwala
12:05 The IRR of a cow (and why Excel lied)
14:51 Bootstrap or raise? The real answer
17:00 Raising from friends & family without ruining relationships
21:02 From ₹80 lakhs to ₹200 crores a month
23:01 The milkman model & month 60 retention
32:34 Final advice: take more risks, there is no downside

Are entrepreneurs born or made? Asish Mohapatra is certain it's the second and he'll tell you plainly he's good at maybe 3 of the 10 things a founder needs.

Asish is the co-founder of OfBusiness and Oxyzo, which he built into one of India's largest B2B commerce and lending businesses. But this conversation with Avnish isn't about the scale, it's about the intellectual honesty underneath it.

In this episode, they get into:

1. How did you get comfortable in your own skin? Was there a trigger?

2. Are entrepreneurs born or made, and how do you decide which skills to build vs. buy?

3. How do you pick a co-founder (and run a company with your spouse)?

4. Why would you never hire someone at their last salary?

5. Do you have to go public and when?

6. A conversation about failing early, being honest about it, and finding the 

🎧 New episodes of Unstarted every Thursday. For founders. By founders.

Chapters

00:00 Cold open
01:30 The boss he fired, the mentor he kept
03:30 The 10% you keep for life
07:30 Are founders born or made?
08:30 Why I got comfortable failing
08:45 Build it or buy it?
15:30 Picking a co-founder (and a wife)
19:00 Why I work 20-hour days
19:35 Baby is six days older than the company
22:45 The vulnerabilities he leads with
28:00 Never hire at their last salary
30:30 Always the next IPO?
35:00 Be yourself. Keep building your superpower.

Can you be a founder without ever founding anything?

Amarjeet Batra has spent 25 years building other people's companies, first Baazee, eBay, OLX, and now Spotify India, and never once thought of himself as an employee. He's what Avnish calls "professionally unstarted": a founder from within.

Avnish and Amarjeet get into the questions most operators never say out loud:  

1. If you have the skills, the confidence, and the network, but not the one big idea, what do you actually do with that? 

2. Is raising a fund a solution, or a responsibility you take on before you've found the problem?

3. Why would you choose 1% of a billion-dollar company over 100% of a ten-million-dollar one? 

4. How do you build a category when ten players already exist and you've arrived last? 

5. When is a difficult problem worth solving, and when does the market simply not care enough to pay?

6. This is a conversation about range over specialisation, ownership without a cap table, and why some of the most entrepreneurial people you'll meet never start a company of their own.

Chapters

00:00  Cold open
01:30 The professionally unstarted founder
02:55 Baazi, the born-again moment
05:10 Why I broke every rule of specialization
08:30 The eBay epiphany: time to do something bigger
09:43 China, and the scale that humbled me
18:26 The power of moving last
20:05  Building the category nobody built
21:15 Is there still a marketplace to win?
22:56 Disruption, distribution, and the AI shift
23:58  Q: How do you actually scale an events business?
26:15 Q: (cont.) Why your event might be the wrong product
28:14 Q: Should we build a place to apprentice under founders?
31:32 How to actually reach a busy operator
34:20 Q: Difficult problem, or one nobody will pay for?
35:09 Problem-first, and the willingness-to-pay test
36:37 Vitamin or painkiller
39:01 Play-front music: a business model flips
41:09 Why success looks overnight

What does it actually take to build AI for 70 crore users?

Vikram sits down with Rahul Chowdhury -  co-founder and CTO of PhonePeto talk about how India's most scaled fintech is approaching AI. Not with hype or a top-down mandate, but with a quiet, deliberate, engineering-first philosophy that started four years ago with a small team focused on making developers happier.

Rahul shares the inside story of PhonePe's AI journey from building their own LLM gateway and Agent Hub, to launching AI search with Microsoft, to betting on on-device models for privacy and cost. And it ends with the biggest idea of all: India's DPI stack has spent a decade making data AI-ready. 

The opportunity now is to use it to build the bank branch of one — truly personalized financial products for every Indian.

If you're a founder, engineer, or product leader trying to understand where India's AI story is really headed, don't miss this.

What you'll learn

🔹 Why PhonePe avoided output metrics in year one of AI and why it worked

🔹 How to build an AI culture without a top-down mandate

🔹 What an LLM gateway is and why every scaled company needs one

🔹 Why on-device models are the right bet for consumer AI in India

🔹 How DPDP will reshape how companies think about AI and data

🔹 Why India's role in global AI is in applied AI — not foundational models

🔹 How DPI × AI creates the opportunity for hyper-personalized financial products

Chapters

0:00 Intro & who is Rahul Chowdhury
02:30 PhonePe's AI journey: tinkerers to transformers
06:00 The DevX team: why developer happiness came first
10:00 Don't rush into AI — the engineering first mindset
14:30 Building the LLM gateway & data stack
18:00 Agent Hub: PhonePe's internal marketplace of agents
22:00 AI Search with Microsoft & on-device models
26:00 Why India needs edge models, not foundational ones
30:00 DPI × AI: the bank branch of one
34:00 Conclusion

Harshil Mathur started Razorpay after quitting the highest-paying job on his campus, a role his whole family had just celebrated, because he walked in on day one and realised he was a guy who wanted to sit and code, not step onto an oil field.

Then he spent a decade away from that: walking into bank after bank getting laughed out of the room, surviving the grind no funding can fast-track, and the night Yes Bank froze with customer money stuck inside it. This is the founder story, lived experience as an edge, why the rejections compounded in his favour, why the grind always comes, and the values that made the hard calls simple.

And then the thing that pulled him back: agentic AI. "It went from being an assistant to an execution engine." Six years after he last wrote real code, Harshil locked himself in a room, asked "if I were to start Razorpay today, how would I build it?" — and rebuilt everything.

The second half is an operator's view of what that shift actually changes: 

1. Why AI magnifies an org's weaknesses instead of fixing them
2. Why an agent with no plan drifts exactly like a company with no plan
3. How Razorpay flipped its leadership hackathon and the bet behind Agent Studio
4. Hosted by Avnish Bajaj with Vikram Vaidyanathan this is a conversation about building, walking away from it, and being pulled back, and what that says about where AI is headed.

Chapters

00:00 Introduction
02:15 Growing up in Jaipur & coding since 6th grade
05:30 IIT Roorkee, SDS Labs & building without permission
10:45 Quitting a $100,000 Schlumberger job in 6 months
14:20 The Facebook comment that sparked Razorpay
18:00 100 banker rejections & how rejections compound
24:10 Getting into YC with zero expectations
35:30 Yes Bank freezes — one decision defines the culture
40:00 Going back to coding after 6 years — AI changes everything
52:00 Rebuilding Razorpay from scratch with AI agents

Follow Z47

Website - https://www.z47.com/
Instagram - https://www.instagram.com/z47.vc/
LinkedIn - https://www.linkedin.com/company/z47-vc/

Anjali Sardana grew up in northern Virginia, studied biology at Georgetown, worked at Bain Capital — and then, without telling her parents, flew to India and founded Pronto: a platform building the world's largest labor organization network, starting with home services.

In this episode of Unstarted, Anjali breaks down how she picked an operations business over a product business (and why), why she sees India's informal labor market as a trillion-dollar opportunity, and the founder mindset that got her through the messy, chaotic, sleep-deprived early days.

She also gets brutally honest about faking confidence, hiring missionaries not mercenaries, and why she thinks most human limitations are completely made up.

Chapters

0:00 Intro — Meet Anjali Sardana
1:20 Growing up in Virginia, studying biology at Georgetown
3:10 The evolution framework that shaped her business thinking
5:00 Product vs. operations vs. distribution — how she chose
8:30 Why India? The labor-market thesis
12:00 Moving to India with zero experience — and hiding it from her parents
15:40 Fake it till you make it: raising a seed round at Bain Capital
19:15 Running pilots, vibe-coding the app, and getting the first bookings
24:00 The Kapil story — recruiting 30 workers in one afternoon
28:00 Operating 24/7 with 5 people, sleeping in shifts
31:30 Building culture: missionaries vs. mercenaries
36:00 Urgency as a core value — actions beget information
39:30 Conviction vs. market signals — how to balance both

India's GPU footprint is on track to grow 40x by 2030, from ~50,000 today to a couple of million. 

That number is bigger than any public forecast. Sharad Sanghi has the unusual standing to make it: he built Netmagic into India's most significant datacenter business, and he's now running Neysa, the only neo cloud in India that Semi Analysis has rated, backed by Blackstone.
In this episode of Intelligent Indians, Rajinder Balaraman and Sharad cover:

1. Why neo clouds exist as a category, and what hyperscalers structurally can't do for one market 

2. The ITQ case study: how to define ROI before infrastructure 

3. The three infra mistakes that quietly cost AI teams 10x their compute spend 

4. Why power, not GPUs, is the real bottleneck, and why 50% of India's data centre capacity sits in one city 

5. What  India's AI Mission could actually unlock in the next phase

If you're building AI infrastructure in India, tracking the space as an investor, or working on policy in the area, this is the operator view. 

From someone whose entire balance sheet depends on getting the call right.

Chapters 

00:00 India's AI Moment The Big Picture
02:00 Welcome Introducing Sharath of Neysa
03:30 How He Built India's First Data Centre with NetMagic
06:00 How ChatGPT Sparked the Idea for Neysa
18:00 India is 2nd Largest AI Consumer 
21:00 50,000 GPUs Today. 2 Million by 2028 
24:30 Neysa vs AWS, GCP, Azure 
28:00 Why Indian Banks Are Early AI Adopters
31:30 Financial Services, Healthcare, Manufacturing 
35:00 PhonePe, Perfios, Hungama - Real AI Use Cases in India
38:30 Why Most AI Projects Stay in Pilot and Never Reach Production
52:30 GPU Obsolescence Risk — How Neysa Manages It
55:00 Healthcare, Education, Agriculture — Where Founders Should Build
58:30 IIT Bombay and the Bharat Gyan Project
1:01:00 Why India Needs to Keep Its AI Talent at Home
1:04:00 Why He Refused to Flip the Company Outside India
1:06:30 What It Takes to Make India the AI Research Capital of the World