Astha Agarwal

A little mortified Ajay Bulusu posted my tweets on "Cab Company CEO" here. Twitter & LI are like Oil & Water - they don't mix well. But to summarize my thoughts, the attempt by the Cab Company CEO to package Cloud and AI as a "nationalistic" venture, is dubious & unworkable. Very soon it will turn into an attempt at regulatory capture in the name of patriotism & national security (he has already called for data localization using the CrowdStrike outage as an excuse - never mind that these have no relation). I have seen Gaurav and Ajay build NextBillion.ai into a viable alternative to Google maps for enterprises. Krutrim is welcome as well - but no special favors because it's 'Indian'. The world of digital products & ideas is global. More innovation lies ahead than behind - and it will come from talented people - not legacy factories or fertile land. Countries that do well in this will be the ones who attract the best innovators. Technologists tend to cluster around these leading innovators. Capital too is globalised and will fund these innovators wherever they are. Digital innovators are nomadic - they only need an internet connection as a tether - & the world's leading and wannabe tech hubs are all competing to attract them. Our benchmark for success should be how many innovators setup companies in India, work from here, patent their IP here, bill their clients from here. Today, most of our innovators incorporate abroad & emigrate at the first chance. News of relaxed entrepreneur visa norms in the US gets a million views and likes. Many of our best engineers are always looking for emigration opportunities. If the Indian Govt. is interested in capturing more of the value of this ecosystem - it should try to align governance to attract this crowd. That means low and predictable regulation, economic and social freedoms, moderate taxes, great livable urban spaces and so on. This is a far far cry from the India of today - with it's urban hells, parasitic taxation levels in certain areas, ridiculous and high friction regulation around global currency movement and securities transactions, and lack of a trustworthy judicial and L&O system. Hucksters step into this lack of an internationally competitive environment - and claim that "protectionism" in the name of "nationalism" is the answer. It is most definitely not. We won't get our own Facebooks and LinkedIns by erecting walls. Such walls will only help give these hucksters' ventures a leg-up - so their otherwise shoddy goods and services are not exposed to international competition. And in fact such walls will only drive further emigration of genuine innovators. To the contrary - we should be tearing down walls like FEMA laws - that harass economic freedom today. We are the kids of 1980's India. We don't want to go back there. We don't want our internet captured by reincarnations of the Bajaj Auto regulatory capture model (the sheer irony). Compete fairly in the global marketplace and win.

238

16 Comments

Post on Benefits of Fasting Over the last few months, I have been studying the benefits of Fasting ( Intermittent and Prolonged ) on overall Health and reversing diabetes. I would like to thank Lalit Kapoor Sir for inspiring me to explore and research this topic. I have compiled the learnings into an AI chatbot accessible at https://lnkd.in/gAYgYE-E This is still work in progress and I would love to get feedback if AI chatbots trained on particular topics are helpful - please feel free to try it out and share your comments on its usefulness. Based on the feedback, I will create AI bots for other topics as well. #fasting #AI #diabetes-reversal

32

Ever had that moment when your engineering team says a feature will take 3 months to build, and you're thinking "Really? 😅" Well, something wild happened yesterday - I and Pranay Gaddam⚡ built an entire YouTube-style platform for Product Management content in just 90 minutes! Not a mockup, not a prototype, but a working product. I know, I know... I couldn't believe it either! 🤯 Here's the fun part: We are not some 10x developers or coding wizard. We just paired up with Replit's AI Agent (think of it as having a super-smart coding buddy who actually listens to what you want 😉). We recorded the whole journey, showing exactly how it works, because let's be honest - as product folks, we need to understand this AI revolution beyond just the buzzwords. What you'll see: • Real-time problem-solving (including the head-scratching moments!) • How AI actually understands and implements product requirements • The future of rapid prototyping (spoiler: it's already here) • A peek into what's possible when AI meets product development The best part? No coding experience needed to follow along. We explain everything in product manager language - no tech jargon overload, promise! If you're a product person trying to navigate the AI landscape (aren't we all?), come geek out with me: [Link in comments] P.S. For all my PM friends who've ever wished for faster development cycles - this one's especially for you! 😊 ---------- 🎓 Learn Product Management - Link in bio #ProductManagement #AI #Innovation #RealTalk #ProductLeadership

13

4 Comments

Are you a Voter or a Vote Bank? As India is going through the general elections, the air is thick with debates – much like trying to dodge your manager on a Friday evening. But amidst the noise of assumptions, logical fallacies, and the infamous 'WhatsApp University,' how do we make informed decisions? Let's start with some common debate pointers vs what data says: 1: Fuel price and inflation skyrocketed just in the last 10 years. Here are some common indicators: Average CPI Inflation: 2004-2009: 9.1% 2009-2014: 9.4% 2014-2019: 3.4% 2019-2024: 5.9% Average food inflation: 2004-2009: 12.3% 2009-2014: 11.9% 2014-2019: 0.7% 2014-2024: 7.1% Lets also look at Petrol Price: 2004: INR 35 2014: INR 72 (>100% up) 2024: INR 100 (~40% up) 2. Economy grew faster in the last 10 years: Let's look at some of the indicators: GDP growth rate: 2004-2014: 6.8% (including 2008-09 global recession) 2014-2024: 5.7% (including black swan events like Covid) GDP per capita: 2004-2014: 50% rise using inflation adjustment, 204% increase at absolute value 2014-2024: 43.6% rise using inflation adjustment, 114% increase at absolute value USD to INR: 2004-2014: INR 43 to INR 60 (~39% drop) 2014-2024: INR 60 to INR 83 (~39% drop) Debt to GDP Ratio: 2014: 67% 2024: 82% Non performing assets: 2014:INR 5 Lakh Crore, Write-off: INR 32000 crore 2024: INR 20 Lakh Crore, Write-off: INR 14.56 lakh crore Foreign Direct Investment: 2005-2014: ~USD 280 Bn 2014-2023: ~USD 596 Bn Forex Reserve: 2004-05: USD 141 Bn (22% of GDP) 2014-15: USD 341 Bn (16.3% of GDP) 2024: USD 642 Bn (17.4% of GDP) All this data has been collected from available sources of Govt. of India, RBI, IMF, World Bank, and media reports of credible media houses. There are more indicators, more perspectives, and more arguments. These data points and trends will definitely help you cut through the noise of debates and be more aware. Data and analytics can be a strong tool to make comparisons of different governments on quantifiable topics. It can equally be used to disguise you from the real fact. My attempt is in no way to hide/disguise but to reveal the truth to the best of my knowledge. You can add your perspective in the comments to enrich it. Why I wrote this post: I generally have an inclination in reading and learning about socio-economic and political affairs in my leisure time. In my professional work, I do data and AI day in and day out. So I thought of using some of my off-hours to do comparisons of governments between 2004-2014 vs. 2014-2024 to make decisions like a voter rather than a vote bank. I will deep dive into other governance parameters to compare 2004-2014 vs. 2014-2024 in the next 30 days. All these will be purely based on numbers, no opinions, no arguments. PS: Nothing in for or against any ideology or party. It will be just my thoughts and analysis in my leisure time. You can share your views in the comments. You can correct the data in comments. #India

47

4 Comments

Waiting at Chaayos, where ordering tea feels like signing up for a home loan. Multiple clicks to sign up, log in, and build a cart—all just to order my chai. I’ll think twice before entering their shop again. Seems like their PM traded growth for vanity metrics, like collecting email IDs. Simplicity is key, not chaos! Maybe it's time to rebrand as ChaaOos! #OverEngineering #UXMatters #ProductManagement #ChaaOos #CustomerExperience #KeepItSimple

20

How to get started with #AI? AI leader Saumya Bhatnagar leverages OpenRead to read at least one research paper daily, continuously expanding her knowledge. Expand your knowledge now: https://lnkd.in/gq__88eS #research #paper #reading

5

Startup life can be chaotic. I spoke with Jeevanshu Narang about the challenges of designing in a fast-moving environment without established processes. Clip from Episode 2 for The Founding Designer series. ----- Full episode: https://lnkd.in/g76UMNzg ----- 1. He explains the work culture difference between a startup and corporate 2. Everyone in satrtup works on everything 3. In the case of Verak, pre-PMF product posed the lack of certain direction. and more ----- About the series: The Founding Designer is a series by UXANDMORE(Manav Madaan) where he invites founding designers and hold in-depth conversations about their experiences of working in a fast-paced environment with no existing design function. Topics include: Challenges faced by designers at early-stages Evolution of design function Design leadership at startups Shift in work quality post funding how to evaluate an early-stage startup and their founders. ----- #startup #business #founder #foundingdesigner #fintech #insurance #cars24 #verak ##finance #uxdesigner #product #uiux

39

As I sit with son on factorisation of polynomials... Example: a-square -5a -6 The trick is to multiply the coefficients of a-square and the constant term. You get -6. And then find the numbers that on multiplication gives -6 but on summation gives -5 Now he ponders on 3 and 2 the more obvious ones. Soon he realises that since the signs need to be opposite, they cannot on addition get you to -5. That 1 is a factor of all numbers and the number itself is then recollected. So you get 1 and 6, one of which needs to be negative. Now you come to an interesting part. Which one is to be kept negative? Things are a bit involved by now and so the attention doesnt come on the obvious point. After some pondering, it took some pointing attention to the fact that to get to -5 on addition, it is the bigger number that needs to be made negative. It was a wow moment, even if it was obvious... As i observed him solving subsequent problems, he started remembering this new rule, which he knew from earlier, but just got reinforced. Point 1 - Mathematics Education Now one thing is what it tells us about Mathematics education. The 100 small things to which attention need to be paid to, to get Maths right. It is cognitively taxing. You need supervisory/facilitating human attention, which is the simplest way to motivate a child to go through with it. It can become fun, but only if you get it right often. And like in all learning if you dont get it right often enough, it can make one lose interest. Point 2 - Neural net learning But to come to my soapbox, the neural networks also seem to be learning a 100 small things. The combination of getting trained where a pattern repeats and reinforces is what makes learning happen. The rule is not what we memorise. The rule is the compression of patterned enumeration. But human beings seem to internalise it, not based on memorising rules. We seem to internalise it based on the repetition of the enumeration and the fact that, now you can jump the step and not have to enumerate at all. The Rule is not what is either Taught or Learnt. The Rule is what you arrive at through a process. The process is the observation of the repetition of the enumeration and then suddenly being able to "see" the pattern. And for the LLM to learn any new Rule, we need to give it the Data which has this Patterned Enumeration. We thought in 2023 that we had some generalised Intelligence. We thought wrong. Anyone spending 1000s of hours with LLMs now knows that the LLM is not able to get into too much of generalisation. Neither are most Human Beings able to do that when they Learn! I will be very curious to know how AGI can be achieved. Because like the Ring in the Lord of the Rings, it will have to be the Ring that rules them all! How will we manage to pull that off? Does the patterned enumeration of Rules, have a Rule? Do we Humans somehow get it? Can we produce Data that makes this happen for Neural nets? #humanlearning #agi

4

1 Comment

As Shreyas Doshi calls it There are 3 levels of product work: 1) Impact 2) Execution 3) Optics Each is important. AND Everyone has a default level they enjoy. I focused mainly on execution-level tasks and underestimated the importance of time allocation and high-impact activities. This approach resulted in delayed promotions. To achieve faster promotions, it's crucial to understand your manager's expectations and the importance of all three types of product work. Align your efforts accordingly. #productmangement #google #promote #amazon #meta

60

3 Comments

In 2014, India's average data consumption per user was just 0.3 GB/month at ₹269/GB. Today, thanks to Modi Govt's push and players like Jio, users consume 30.3 GB/month at ₹10-13/GB. Remember when we rationed data usage to stretch it for a month? Now, everyone's a legal expert, taxation guru, election analyst, military strategist and more! 😆 We've come a long way from just worrying about onion and tomato prices to debating complex tax issues like Indexation. Cheers to progress and thanks to Modi Govt for making us a more developed nation!

681

57 Comments

In 2008, two UX researchers from Google visited India. This is what they discovered... Indians didn’t navigate using street names. Instead, they relied on landmarks—temples 🛕, Clocktower, or even that one famous “big banyan tree” 🌳. Instead of forcing global navigation norms, Google asked: How do Indians Navigate? The answer revealed the root causes: a lack of reliable street signage 🚫, cultural habits 🌏, and chaotic infrastructure 🛣️. From this foundation, they reimagined Google Maps Platform for India: 1️⃣ Landmark-based navigation: Instead of "Turn left at the main street," directions like “Turn left at Ganpati Temple” 🛕 felt more intuitive 2️⃣ Metro booking integration: 🚇 Simplifying fragmented transit into a unified platform 3️⃣ Road-specific nudges: ⚠️ Alerts like “Take flyover” or “Avoid narrow road” for better routing 4️⃣ Plus Codes: 📍 Providing an address where there was none, bridging urban and rural gaps 5️⃣ Multilingual support: 🌐 Navigating in up to 9 Indian languages to make maps accessible for all (Including auto drivers) This wasn’t just localization—it was first principles in action, solving deep-seated problems instead of patching symptoms. 🤕 As product managers, we often inherit constraints that seem immovable. But true innovation demands we strip down assumptions, ask the why ❓, and design solutions that feel inevitable. What’s one challenge in your product that needs a first-principles approach?💡 #productmanager #productmanagement #google

29

I don't remember the exact day but it happened last year on a usual office day. There was a new leadership change which happened recently due to which everyone (product and tech) was under tremendous stress and the air the office around had been becoming toxic. - Toxic with disrespectful comments and feedback - Toxic with natural fear and resentment developing among the team members - Toxic with random uncontrollable dreams about teams being fired and disrespect in front of whole office! On this day, I went to the office, did some work, and had a chat with my colleagues as usual. While going back home, I got a little late and Bengaluru traffic as we all know is a brutal demon which fills you with frustration. I got overwhelmed with the negative thoughts about office toxicity, traffic toxicity, life toxicity, etc. and started questioning my life choices while I was stuck in traffic. When I reached home, I went to my bae, hugged her and started weeping and did not stop for 7-10 minutes. She just stroked my hair trying to calm me down before asking the reason. Crying out loud was one of the best therapy I realised which might stop you from killing yourself. Having someone who understands your pain and care about you helps you let out the negativity in your head and around you. People are killing themselves, getting sick, and delusional due to work pressure, toxic environment and what not is becoming a norm these days especially in India. I understand that to grow fast you need to work very very hard and out in long long hours! But, with that, Do create a support system, a safety net which protects you if you fall emotionally, financially or Physically! If you are struggling emotionally, please reach out to your loved ones, don't feel embarrassed, it's okay to fall, it's totally fine to feel emotionally overwhelmed. After all, we are humans, we are fragile. #Struggle #Emotional #Distress #CryOutLoud

80

2 Comments

Yikes. I had no idea. Many Indians pre booked Tesla by paying 1000$ in 2016 (which is a pretty decent sum in India ) . With no tesla in sight for 8 years , they are now struggling to get refunds Ps: if they had put 1000$ in Nifty , they would likely have 3000$. Had to add this comparison because people keep talking about investing in company stock rather than enjoying the product, this is one such rare case where this is correct 😬 https://lnkd.in/gqYtQEuJ

17

2 Comments

AI Self Healing is live AQUMEN LABS. Over the past month, the team has been working hard to reduce costs incurred by AI during automation. AI self-healing helped us in : 1. Reducing test writing and automation by 90% - We wrote a test for login and completing an FTUE of a game in just 2 minutes [Faster than other low code tools] and not to mention the Game is in UNITY [Yes we support Unity :) ]. 2. Reduce the Maintenance of the test script by 80%. The test will be healed when you select the Heal checkbox on your test step. 3. We have successfully implemented a Hybrid approach, where our framework is intelligent enough to decide whether to use AI, which helped us in reducing the cost. For cost reference: MultiOn [an agentic AI] uses 0.04$ per step and AQUMEN LABS uses 0.03$ per step executed and that too in an intelligent way :) This is purely a Vision-Based Multi-Model approach we do not use any locators to heal. If you and your team are exploring an AI approach in testing, we would be happy to showcase it. Schedule a call: https://lnkd.in/gmu62wzV Note: Used Multion Example as it is one of the best agentic AI tools that we follow and take inspiration from. Multion and AQUMEN LABS are two totally different tools for solving different use cases.

5

I just completed Shreyas Doshi's Product Sense course and I am still so overwhelmed with the exceptional experience. Initially, I was hesitant about the time commitment and whether the course would truly deliver value beyond what I could find in other resources. However, by the end of the very first day, I knew this was a unique and invaluable investment in my PM career. The concepts he covers are profound and advanced, going far beyond surface-level product management techniques. Unlike typical PM books or courses, this program challenges you to deeply introspect and reshape your entire product mindset.The curriculum forces you to confront your own biases, assumptions, and mental models in a way that I've never experienced before. By the end of the 5-day program, my product lens had been completely transformed. I would highly recommend it to any product manager looking to create winning products.

30

if you know, you know Vinod Khosla of Khosla Ventures. nerd. founder. wise. man. been there done that sv. wisdom. it's missing. let's face it. it's always been missing over time. for.ever. so it's a blessing truly to have an o.g. that keeps it for reals. yes we are a fan. much like Jensen Huang Fei-Fei Li ilya sutskever ...and other good people in sv. they are the ones who make this place so special. indeed they do. here https://lnkd.in/exWqey9K we have sam i am that's all that i am sam (alt)man having a 1-1 with the man. this is a must watch to provide a foundational (no llm required here) mindset as to how sv is at it's best (and worse). caution. watch out for the truth bombs. vinod is (in)famous for this style and we <love> him for it. thank g-d he exists. if you meet the dbags in sf/bayarea your experience you will be quickly lonely, dazed and confused. best to listen to these wise words from vinod and from that determine your spot in the sv landscape. with an insight on the sv playbook from a master. well. need i say more. it's about the very spirit of sv that makes it so special. and this is indeed embodied in a few who are masters. deserve our attention. yeah. yeah. i know attention is all you need. although what happens if your attention is being distracted from building. being in the company of those who will join your mission ala brian chesky of Airbnb. missionaries. search for the good missionaries in sv. keep away and keep in mind of the mercenaries. they are there. pg. he's another gift (as opposed to the grifters) paul graham wrote about this too https://lnkd.in/eEYmQwqZ - being a chesky-esque missionary. these are good things to follow on the path of achieving escape velocity as an (ai) startup. oh. hollywood has entered the ranks of sv. yes it has. so there is a bit of performance art that is required. not enough that you've created something that's novel (truly, novel). also requires a bit of showmanship. audacious in fact. it's a well known sentiment that audacious ideas motivate people. sun microsystems. audacious. juniper. audacious. OpenAI. vinod as an early investor. audacious. Vinod Khosla was born audacious. so fitting he found a place in this world where many have benefitted from this sentiment. perhaps years ago a bit on the abrasive side. now with those white hairs. the elder statesman. non-diplomat. rules, the roost of that which allows for others to flourish. note if you manage to thread the needle of vinod means that it's a very very very good beginning. nothing is guaranteed of course. as my lawyer 'friends' like to say. * buyer beware. be aware. Vinod Khosla is. are you (ai)? are you?? perhaps yes. perhaps, non. < . >

India Lost the Test: Here’s the Product Management Fail You’re About to Make 😱🏏 India got completely obliterated in the 2nd Adelaide Test. And because I’m a product manager, I’m going to milk this failure for some “valuable” lessons. Buckle up. 1. Overconfidence = Disaster India thought they had it in the bag. Spoiler: They didn’t. Much like launching a feature you think is going to change the world. It won’t. Keep it simple. 🧐 2. Stick to What Works (Or Don’t, and Watch Your Product Fail) India kept trying to bowl the same way despite the conditions. If your product isn’t working, just change the damn strategy. If not, you’re as good as India’s batting lineup today—useless. 🏃‍♂️ 3. The Art of Accepting Failure India’s team had no idea how to handle the collapse. You’re gonna screw up too. Get over it. Accept it, learn, and move on before the stakeholders start calling for your head. 😬 4. Pivot Fast or Die Slow India stuck to their game plan like it was their baby. Guess what? It wasn’t. Like your product roadmap—if it’s clearly going nowhere, stop wasting time. Pivot before it’s too late. ⏳ 5. Resilience is Key (Or Just Lie About It) India’s loss doesn’t mean the game’s over—except if you’re in product, where it often does. But hey, if you fail, just say you’re "learning" and pretend like you have a comeback strategy. Classic. 😜 So yeah, India lost, but I just turned it into a crash course on product management. See? Everything can be a teachable moment—if you're not too busy crying over your last product fail. #ProductManagement #CricketFail #LifeLessons #AdelaideTest #PivotOrDie

4