Keshav Malpani

🚨 Struggling with Low-Frequency Products? Here’s the Playbook I Wish I Had 🚨 For 15+ years, I’ve been managing products that people don’t use every day and only on episodic basis. And let me tell you—it’s tough. No guide. No clear answers to questions like: 👉 How do you measure product-market fit? 👉 How do you optimize acquisition costs when usage is infrequent? 👉 How do you actually drive retention? The frustration led me to create ICED Theory—a framework so intuitive and logical, it became part of Reforge’s course material. Now, for the first time ever, I’m hosting LIVE virtual classes to teach the ICED framework and help you grow and manage low-frequency products. 📅 Two intense days of interactive discussions. 💡 Deep dives into strategies that work. 🎉 50% launch discount for my first cohort. 👉 Ready to transform your approach to low-frequency products? Enroll now and secure your spot: https://lnkd.in/gHwUsGiJ Spots are limited, so don’t wait! Let’s crack the code together. 🚀 #lowfrequencyproducts #workshop #productmanagement #training #cohort

34

1 Comment

🚀Hi Folks, I'm Taking on the Product #Teardown_Challenge along with Arindam Goswami! 🚀 As part of an exciting challenge by Product Space, I’m stepping into the role of Senior VP of Product at #Zepto Quick Commerce! Zepto’s already crushing it with 10-minute deliveries across 30 cities in India, but there’s so much more potential to unlock! 💡 Here’s the mission: ➡️ Increase Zepto’s Average Order Value (AOV) to ₹600 🛒 ➡️ Boost product discovery and engagement to encourage bigger, more frequent orders 🔍 ➡️ Elevate the purchase experience to keep customers coming back for more 🔄 And I need YOUR help to make it happen! 🙌 🎯 Your insights on quick commerce platforms like Zepto, Blinkit, and Swiggy Instamart are key to shaping these strategies. I’ve put together a short survey, and your feedback will be crucial in tackling this challenge! 💥 Thanks for your support—let’s crack this together! 🙌 Link to the form - https://lnkd.in/g5v2y4wR #ProductTeardown #QuickCommerce #ProductManagement #ProductSpace

10

2 Comments

I'm thrilled to be speaking about "Apache XTable (Incubating): Interoperability across Open Table Formats" at The Fifth Elephant 2024 annual conference tomorrow at BIC, Bangalore organized by Hasgeek In this talk you will learn about: 1. What are Table Formats? Why do we need Inter-Operability ? 2. A deep dive on XTable and problems that can be solved using XTable. 3. A demo of XTable in action. 4. The roadmap for the OSS project. You can check out and give a star to the OSS repository for the project here: https://lnkd.in/g53VWyU4 🎟️ Get your tickets and join us in-person: http://has.gy/sSCX P.S First 5 people to buy a ticket to the conference using the above link get a 10% discount! So go grab your ticket fast! #Hasgeek #AI #TheFifthElephant #MachineLearning #DataScience #BigData

85

How do you build infra successfully at a startup? 1. Build something quick and functional, but inefficient and unreliable 2. Prove that people care that your product is expensive and unreliable 3. Optimize Checkout this awesome blog by one of my closest collaborators at Statsig, Tore H., on how we got Statsig up and running successfully.

15

1 Comment

𝐏𝐫𝐨𝐣𝐞𝐜𝐭 𝐓𝐡𝐢𝐧𝐤𝐢𝐧𝐠 vs. 𝐏𝐫𝐨𝐝𝐮𝐜𝐭 𝐓𝐡𝐢𝐧𝐤𝐢𝐧𝐠: What Are They? It's worth validating if you think you know. About three years ago, my product mentor pointed out that we often 𝐛𝐞𝐡𝐚𝐯𝐞 𝐦𝐨𝐫𝐞 𝐥𝐢𝐤𝐞 𝐩𝐫𝐨𝐣𝐞𝐜𝐭 𝐦𝐚𝐧𝐚𝐠𝐞𝐫𝐬. This sparked a question within me, and instead of asking my mentor, I decided to learn independently (call it a mistake). I explored numerous books, blogs, and videos, delving into user journeys, understanding user problems, and thinking creatively about user needs. I jumped between various concepts, believing them to be product thinking 𝐛𝐮𝐭 𝐧𝐨𝐧𝐞 𝐬𝐚𝐭𝐢𝐬𝐟𝐢𝐞𝐝 𝐦𝐞.   Then, one day, I had the chance to listen to Shreyas Doshi. It is a free masterclass, and that's where I found my answer. Here are the five key principles from his workshop: Suspend the project-thinking mindset - Take your 𝐩𝐫𝐨𝐣𝐞𝐜𝐭 𝐭𝐡𝐢𝐧𝐤𝐢𝐧𝐠 𝐡𝐚𝐭 𝐨𝐟𝐟. Put the product thinking hat on Prioritize your real goals - Ask 𝐖𝐡𝐲? And 𝐒𝐨 𝐰𝐡𝐚𝐭? Often. What 𝐞𝐟𝐟𝐞𝐜𝐭𝐬 𝐝𝐨 𝐲𝐨𝐮 𝐰𝐚𝐧𝐭 𝐭𝐨 𝐜𝐫𝐞𝐚𝐭𝐞 on your user? Understand your user needs: Pay particular attention to 𝐨𝐛𝐣𝐞𝐜𝐭𝐢𝐨𝐧 𝐚𝐧𝐝 𝐟𝐫𝐢𝐜𝐭𝐢𝐨𝐧 𝐩𝐨𝐢𝐧𝐭𝐬. Look for 𝐮𝐧𝐞𝐱𝐩𝐫𝐞𝐬𝐬𝐞𝐝 𝐧𝐞𝐞𝐝𝐬. Generate options: Don’t be afraid of 𝐛𝐢𝐠 𝐢𝐝𝐞𝐚𝐬. Embrace 𝐜𝐫𝐞𝐚𝐭𝐢𝐯𝐢𝐭𝐲 𝐚𝐧𝐝 𝐝𝐢𝐟𝐟𝐞𝐫𝐞𝐧𝐭𝐢𝐚𝐭𝐢𝐨𝐧 Simulate: 𝐕𝐢𝐬𝐮𝐚𝐥𝐢𝐳𝐞 how each option will 𝐩𝐥𝐚𝐲 𝐨𝐮𝐭. Then 𝐫𝐞𝐩𝐞𝐚𝐭 until you converge.

8

Excited to share some of our work on building an in-house text2sql tool. We built Hermes, our gen AI based text2sql workflow, to increase data accessibility and as a small but significant step towards self-serve analytics. Rutvik Vijjali the lead data scientist and Amaresh Marripudi the Sr PM on this project share our learnings and evolution. At the outset it is a 'straightforward' RAG based implementation using commercial LLM APIs but the devil is in the details, practical considerations and the edge cases we uncovered and tried to solve along the way. For example, many commercial solutions seem to imply that you can throw the kitchen sink of all your (100s of) tables and (1000s of) columns at their text2sql engines and expect to receive accurate SQL queries. Except for some very simple use-cases, this rarely works in a real-world business setting. Instead, we compartmentalize our business/ functions into logical units and instantiate separate text2sql engines for each. This design choice gave us more benefits (for ex-- more accurate SQL generation, faster onboarding due to relatively smaller amount of metadata & curation needed) than trade-offs (for ex-- can't yet generate queries that cut across businesses). Similarly, figuring how to make use of hundreds of thousands of queries people have executed over the years, was another unlock. While we have a long way to go, the adoption, repeat-usage and feedback stats have been strong. We have a host of updates (for ex-- augmenting with a sql2text built on historical query corpus, adding explainability, moving to agentic flows) and features planned in the coming months. #swiggy https://lnkd.in/g7q89YEY

670

37 Comments

I am not sure Google’s primary duty should be to shareholders, including mass firings to pay for GPUs. As a shareholder, I am thankful to be able to own a few shares of Google in my retirement account. But Delaware Supreme Court case law serves unjustly as a faux federal loophole to maximize profits above all else, enabling amoral fiduciary duty to conventional shareholders. And it seems that employees are shareholders - should one class of shareholders be favored over another? It seems like the law needs to include stakeholders like the economy in which a company arises and the communities in which it does business. If a given government has a shrinking middle class, is it not true that eventually the corporate tax base will not be enough to pay for adequate national defense or safety nets? I happen to be pro-Google, especially because of their contribution to help so many businesses utilize the trackable ROI of Google Ads, and for many other reasons. As far as I know they obey the law - and it seems like the law needs to change. I also happen to believe it is good for the world to spread economic opportunity and jobs. But what makes up for all the money and tax these core Google employees were spending on the American economy, or the people who will lose out on a job opportunity because a Xoogler applied? Housing costs in SF are insane, what about their mortgages? Maybe they are millionaires, maybe not. I kind of doubt Google fired AI engineers, but maybe instead of hiring domestically they will hire wherever the AI labor is cheapest. I happen to be pro-immigration and my ancestors were immigrants. I don’t pretend to presume the right balance of domestic and offshore workers. I have been very pleased to collaborate with people and pay them out of pocket and be able to afford it because of the labor discount. But I am not a hugely profitable corporation either. I tend to believe in unfettered capitalism and that too much tax and regulation can stymie an economy and hurt everyone. Yet an economy needs to be sustainable, and not just for conventional shareholders. I am not sure there is a check and balance to the “duty to shareholders”. Public outcry used to be one, but not anymore, not after the massive tech hiring and layoffs. Maybe it is just competition. But it doesn’t seem biomimetic, as in natural balance between various forces in a biological system. If an economy was a body and large corporations were organs in the body, if money was water and employees are like blood, then the corporate organs seem to be growing abnormally and taking more moisture out of the system. In a normal body the organs would have limited growth but in this body the corporate organs are growing larger, while the rest of the body slowly dries out, and blood clots form from employees who are no longer needed. The corporate organs are less reliant on employees and are now making their own blood and selling it.

11

3 Comments

Many of ML practitioners might have heard the term "Extending context length" for LLMs. In case if you want to dive deeper into this topic along with its code implementation, give this blog a read. I have described about Position Interpolation, NTK aware and Dynamic NTK method, which are widely used for extending context length.

69

HIRING FOR PRODUCT MANAGER 🔥🔥🔥 ​​We're looking ~3 years-experience Product Manager at SuperApp (YC W18). Just like Lego boxes that fit on top of one another, we are looking out for someone who has experience that can be plugged together, customized and deployed anywhere. Someone who is passionate about product features, customer needs, market research, and taking up full ownership of business goals. . 🙍Product Manager ~3 years experience. Office: Alam Sutera (Hybrid) https://lnkd.in/dEQDjUXr (apply here) . Expectations in the first week ✅ Focus: Absorb and Learn ✅ Expectations: The key thing is to actively listen and learn as much as possible. This includes: - Meeting new people and understanding the company culture - Getting set up with your workspace and tools - Learning about the product and its functionalities - Introducing yourself to key stakeholders and understanding their expectations . Expectations in the first month ✅ Focus: Deep Dive and Understand ✅ Expectations: Now that you have a basic foundation, you should delve deeper into the product and its ecosystem: - Understand the customer base and their needs through various methods (customer calls, surveys, user research) - Learn about the competitive landscape and how your product positions itself - Familiarize yourself with the product roadmap and development process - Begin building relationships with your team members (engineers, designers, marketing) . Expectations in the first three months ✅ Focus: Apply and Contribute ✅ Expectations: With a solid understanding of the product and your role, you can start applying your knowledge and contributing more: - Participate in product roadmap discussions and suggest improvements - Develop a data-driven approach to product decisions using analytics - Start working on assigned projects and contribute to the product development cycle - Communicate effectively with stakeholders and keep them informed . Hiring Stages Round 1 : HR Screening Round 2 : Interview with Head of Product and/or CPO. Round 3 : HR Interview (offering). . Why work with us At Aplikasi Super, you have the opportunity to ✅ Flexible working hours at Alam Sutera or Surabaya. ✅ A flat and transparent culture. ✅ Enjoy a culture with opportunities for exponential growth and learning with 7 squads (cross-function). ✅ Laptop Ownership Program and Learning/Course Wallet after finishing probation. . How to apply If you're interested in the role, click the link above. If you know someone talented who might be a good fit, please let him know by mention his/her name or "like" this post so we can help others too! If you're curious what's SUPER Culture, you can check my other posts or SuperApp (YC W18). For any queries you have, DM me! . Sharing is caring. We're SUPER, we're together! ~~~ #hiring #opportunity #JadiSuper #productmanager #productmanagement #hiringproduct #job #jobalert #projectmanager #projectmanagement

42

4 Comments

Recently, I took Shreyas Doshi's Product Sense course. Undeniably, it has been the most transformative cohort-based course I've taken, which massively over-delivers on its promise. Shreyas is the kind of instructor who will nudge you to think deeply and distill something as ambiguous and complex as "product sense". As a Product Manager, I've often struggled with digesting conventional wisdom propagated as core PM skills. This course brilliantly helps you overcome these limiting notions and uncover the truth of product thinking. Highly recommend this course for any PM or anyone passionate about building exceptional products. It fundamentally rewires your approach to product management with the added perk of being in a highly accomplished cohort of co-learners.

26

1 Comment

Join us for a conversation with Peggy Rayzis on building a "Developer Intent led" GTM motion! We explore the key drivers that led to the wild success of Apollo Graph QL... followed by an AMA Link to register in comments. Feel free to share any questions

7

2 Comments

Had a great time at the Uber Gen AI Meetup in Bangalore! As I'm starting my Gen AI learning journey, it was amazing to see how experts are using AI to solve real problems. Following are some of the highlights from the event: - Jayaram Valliyur set the stage with their welcome and keynote on Uber's advancements in GenAI - Vivasvat Keswani introduced Uber's Michelangelo: Platform to build, deploy, and operate machine learning/ GenAI solutions at Uber’s scale - Sameera Poduri showed us how Uber is using Generative AI to make customer service better. - Varsha Mahadevan from Coinbase discussed the importance of keeping AI safe and secure in applications, focusing on prompt injection and implementing guardrails to prevent misuse. - Lokesh Agarwal from Kognitos discussed how AI could change our future - Pratyush Kumar from Sarvam AI emphasized the importance of using Indian languages in AI, making it more inclusive and also discussed building a full stack GenAI platform - Rush Tehrani from Uber focused on scaling ML infrastructure, deployment and serving of LLMs at Scale - Manyata Goyal and Geeti Oberoi introduced us to Uber’s tool, uLabel, which is Uber's labeling framework / data annotation solution that helps with organising data. - Ayush Garg, Chanakya Malireddy, Radhesh Sarma, and Georgy Mathew brought to light various advanced techniques like RAG-based chatbots, dynamic few-shot prompting, auto evaluation framework, showcasing Uber's leadership in applying AI to solve complex problems. - Smit Shah spoke on extracting actionable insights from LLMs, another testament to the practical uses of AI in business operations. - Ajeet Ganga explained how they're leveraging Gen AI to help manage Uber's workforce better - Panel discussion on "Unlocking the Value of Gen AI" moderated by Sameera Poduri featured Rohit Chatter from Walmart, @Vivasvat Keswani from Uber and Saurabh Dash from Cohere, along with Pratyush Kumar from Sarvam AI, offering diverse perspectives on the integration and impact of AI technologies across various sectors. Grateful to have been part of an event that showcased groundbreaking discussions on Gen AI. Big Shoutout to all the hosts at Uber for organising such a well-coordinated event. Looking forward to more such events! #UberGenAIMeetup #AI #Learning

68

2 Comments

I was in cypher a few years ago and had an interesting encounter with a travel booking platform leader who boasted about using cutting-edge AI like RL. However, all I received from that app were incessant push notifications throughout the day! 😂 Using advanced AI should prioritize accuracy in delivering what users need, rather than bombarding them with irrelevant notifications. Let's leave spamming in the Stone Age and focus on leveraging technology effectively. #AI #Technology #UserExperience 🧐

14

1 Comment

Ran a quick test on the new #OpenAI #o1mini to solve IIT-JEE Advanced 2024 maths questions. It accurately solved 13 out of 17 questions with an impressive accuracy of ~76%. Quick thoughts: 1. The hidden chain of thought is quite detailed and extensive. Breaks the problem, solves and verifies and iterates multiple times. This might increase the output token cost by a big margin. 2. Seems there are tools integrated for calculations. Much needed as #gpt4o does a bad job at it. 3. Latency is quite high. Some questions took more than 50 seconds to get answered. This needs to come down for production user-facing use cases. 4. Hallucination still exists. Check the 1st question in the link, it had the correct step-by-step reasoning but hallucinated at generating the final answer. 5. Overdoing self-thought. Saw a couple of instances where it did generate correct answers in the process but eventually made wrong answers. Would give a clearer picture once the rate limits are increased and benchmarking is done on a bigger sample. Let me know any interesting findings and thoughts 😀 https://lnkd.in/giSCezyG #GenAI #OpenAI #PhysicsWallah

139

5 Comments

❝ How do you taste your Wine? With the right Glass, served from the right Bottle. How do you utilize your LLM? With the right Task, served from the right Compute environment. ❞ If the quote above caught your interest, check out my blog post that was featured on the Toyota Connected India Medium channel last week: Leveraging LLMs in Cloud - 🔗 https://lnkd.in/gfHsdevH --- I hope this blog adds a bit of amusement for the seasoned AI/ML engineers 🧠👨‍💻 with the carefully crafted images and perhaps even the silly wine analogies 😉. Whether you're a cloud engineer ☁️ 👨‍💻, a product owner of cloud-based apps, or someone new to Generative AI in the cloud, this blog offers a detailed, beginner-friendly look into the selection and deployment of LLMs in cloud environments. --- A heartfelt thanks 🙏 to my colleagues at Toyota Connected India for taking the time to review, shape this piece and even share a creative writeup like below in the original post 🫡 👇 🤔 Curious to dive in? Click the link below to explore the blog on Medium! 🌐 🔗 https://lnkd.in/gfHsdevH

18

Perhaps the most critical thing in a Retrieval Augmented Generation (RAG) system is the retrieval or searching strategy. Poor retrieval will lead to irrelevant documents going to the LLM leading to irrelevant generated answers or worse hallucinations. Is cosine similarity everything? Do check out below illustration which talks about four of the most common strategies for playing around with retrieval strategies in RAG systems: 1. Similarity-based retrieval: This is based on embedding similarity, you can play around with different metrics like cosine, maximal margin relevance etc. 2. Similarity with Threshold retrieval: Useful where we cap the similarity threshold and can help in removing irrelevant document but can drop recall 3. Multi-query retrieval: Make the LLM generate variations of your query and get documents based on these multi-versions of your initial query, often useful to get more recall or coverage of documents 4. Reranker retrieval: Use a model specialised in ranking documents based on search relevance Often you can combine some of these techniques and even use ensemble retrievers. Do explore them and go beyond just cosine similarity!

17

Discover how AI is transforming India's MSMEs! With 94% recognizing AI’s growth potential and 87% confident in its productivity benefits, the future looks bright. The majority advocate for accessible AI technologies and training programs, while nearly half emphasize tailored use cases. Explore these insights and more in our latest article. #AIFuture #Meta

3

Discover how AI is transforming India's MSMEs! With 94% recognizing AI’s growth potential and 87% confident in its productivity benefits, the future looks bright. The majority advocate for accessible AI technologies and training programs, while nearly half emphasize tailored use cases. Explore these insights and more in our latest article. #AIFuture #Meta

3

Read this anecdote from someone interviewing for Spotify SWE in Infra team. He was asked to do a "Live root-cause analysis of a failure which previously occurred in Production at Spotify" How the interview worked: You explain your thought process (check connectivity between System A and B via ping, DNS name resolution, look at open ports, etc...) and they tell u the result like "ping returned 'Name or service not known'". It's like traversing through nodes of a decision tree and getting new insights at every checkpoint, and then deciding the next step! Doesn't it sound interesting? What are some interviews that you really enjoyed? #observability #monitoring #engineering #devops #infra

38

2 Comments