Sabhya Verma

CREATIONAL DESIGN PATTERNS We've already explored the Singleton Pattern... Now it’s time to level up with the Factory Design Pattern 🏭 What is Factory Design Pattern? It’s like ordering food from Zomato 🍔 You don’t care how it’s made — you just say what you want, and it shows up. In code terms? You ask the factory for an object, which gives you the right one — no manual object creation in your main logic. 💡 Real-Life Analogy: Imagine a ShapeFactory — you tell it "circle" or "rectangle" and it hands you the right shape object. No new Circle() or new Rectangle() spam all over your code. Just clean, modular, and scalable architecture. ✅ Why Factory Pattern?  • Avoids tight coupling 💥  • Easier to scale 🚀  • Cleaner code structure 🧼 ✌️ Next time someone says “Creational Pattern,” hit ’em with the Factory! Follow 🤝 : Prateek Goel #lld #hld #systemDesign #factoryPattern #learning #faang

82

5 Comments

Most startups aren't getting the most out of their contractors. Having seen dozens of startups hire with Bolto, I wrote a guide to hiring: I'll explain through the example of Sachin Singh Lodhi, our newest engineer (who has been a 10x dev for us). We hired Sachin from the Bolto platform. He hadn't spent time at a US startup before, so we knew we had to be diligent with onboarding. Here's how we did it: 1/ Allow space for learning communication norms. They may not be used to slack. They may not be used to your meeting cadence. They may not know how you do project management. Assume 0 prior knowledge and teach everything. 2/ Don't expect immediate value-add on day 1 An early Bolto customer nearly fired an engineer because they thought they weren't doing anything. They turned out to be amazing and have been with the team for over a year – there was just some necessary onboarding time. Onboard and ease contractors into work properly. 3/ Set them up to win Involve your hires in important syncs. Put them in important slack channels to provide context. Share the wins with them. It keeps motivation high and allows for alignment towards common goals. Overall, what's most important is to treat your contractors like your teammates. They are a part of your team, so value them that way.

30

5 Comments

Every few months, I hear the same statement echoing in founder circles: “Building product is easy now. AI can code it for you. The real game is only sales.” This is one of the biggest illusions in today’s startup ecosystem. Why this mindset is dangerous: 1️⃣ UX is not code AI can generate a login page in seconds. But it can’t design a flow where a first-time user instantly gets the value. Most “AI-built startups” fail here. 2️⃣ MVP ≠ GitHub repo An MVP isn’t just stitched features. It’s the minimum complete experience that solves a pain point. Many AI-driven products lack even that. 3️⃣ Technical debt kills scale Shipping fast is easy. Scaling clean is hard. AI codebases often collapse once real users and data arrive. 4️⃣ Problem–Market Fit still rules No matter how much code you ship, if you’re solving the wrong problem, nothing saves you. Startups don’t fail from lack of features. They fail from lack of relevance. The truth every founder needs to remember: - AI lowers the barrier to entry. - But it doesn’t lower the bar for success. - Startups die not because they couldn’t code, but because they couldn’t create something users love and trust. Closing punch: So yes, AI coders are a gift. But they don’t change the fundamentals: 🚀 Execution is still about users, experience, and solving real problems. Code is just the easy part. Drop your thoughts below ⬇️ — I’d love to hear how you see it. #Startups #AI #ProductManagement #UXDesign #Entrepreneurship #FounderMindset #Tech

14

2 Comments

Everyone's pointing fingers at Soham Parekh for working multiple startup gigs. Let's cut the BS and address the real issue: The companies that let it happen. Several YC-backed startups, with solid funding and strong pedigrees, still failed to notice someone moonlighting across 3-4 teams. That's not a Soham problem. That's a leadership problem. I've seen this sh*t happen before. It happened to me at Nolte. Each time, it forced us to level up: → Improve hiring processes → Clear deliverables and expectations → Implement weekly check-ins that actually matter Look, if someone can work multiple jobs and still deliver results - that's a different conversation. But if they're hurting outcomes and going unnoticed for months? That's 100% on leadership. Don't just blame the engineer. Fix your systems. Build real visibility into what your team is actually doing. We did - and it saved our asses more than once.

36

44 Comments

OpenAI is paying $500K+ for AI PMs. Apple and Shopify are paying $400K+. Here’s how to convert your resume to showcase your experience as an AI Product Manager 👇 1️⃣ Update your headline & summary ❌ “Product Manager with 7 years in SaaS” ✅ “Product Manager building data-driven products with AI/ML” Use keywords like LLMs, embeddings, GenAI, prompt engineering, agentic workflows, model evaluation. 2️⃣ Highlight AI/ML literacy (you don’t need to be a data scientist) Add a Skills section with: ✅ Concepts: model lifecycle, hallucinations, bias, precision/recall ✅ Tools: OpenAI API, LangChain, CrewAI, vector DBs ✅ Data: SQL, A/B testing, analytics 3️⃣ Reframe your experience bullets ❌ “Owned roadmap for marketplace features” ✅ “Designed AI-powered catalog sync that automated 70% of manual SKU entry” Make every line show AI-driven impact. 4️⃣ Quantify results ✔️ Reduced onboarding time from 3 days → 3 hours using AI prompts ✔️ Improved recommendation accuracy by 25% through ML evaluation 5️⃣ Showcase side projects Your chatbot, workflow automation, or RAG prototype? Add it. Hiring managers care more about AI fluency + initiative than a fancy title. P.S. - An AI PM isn’t expected to train models. They’re expected to spot high-ROI AI use cases, frame trade-offs, and ship responsibly. P.P.S. - If you’re already a PM, you’re halfway there. All you need is the resume narrative that proves it. #productmanagement #aipm #resumetips

46

8 Comments

Why is CRED building something like Garage? At first glance, it looks like a side experiment. But if you zoom out, it fits neatly into CRED’s larger playbook. 1. Revenue diversification (Obvious one) CRED has already maxed out credit card bill payments. Garage unlocks a new set of revenue streams: insurance, auto loans, servicing. Each car on the platform is a monetizable node just like each credit card was in their early days. 2. Customer data depth (car ownership = asset profile, better risk models) Owning a car says a lot about financial standing and spending power. By tracking the cars you drive, their age, and its insurance cycle, CRED strengthens its ability to score customers, target credit offers, and cross-sell with precision. 3. Convenience over cost CRED’s premium audience often qualifies for cheaper personal loans at HDFC or ICICI. Yet many still borrow short-term loans from CRED — even at higher interest rates. Why? Because convenience and speed outweigh marginal cost. The data backs this: CRED’s loan book has crossed ₹19,000 crore, with repeat loan disbursement rising from 32% in FY22 to 57% in FY25. This shows customers are returning for the ease. Garage extends this play: by knowing your car’s value, usage, and liabilities, CRED can underwrite larger loans faster. It’s not competing on cheapest credit, but on frictionless access.

24

Getting into YC is easy. Only ~2% of applicants get in. Indian Railways? Try 0.3%. In the last drive, ~30 million people applied for 90,000 jobs. Odds can be 1,800 to 1. Six years of study is not unusual. Here’s the key difference: A YC application takes days. The building takes years. Years in which you’re creating something of value, whether YC accepts you or not. Exam prep takes years of cramming obscure facts. If you miss the cut‑off, you might emerge a decade later with no job and no work experience. Invest your best years in building something that compounds: a product, a company, and a skill set no one can take from you.

16

7 Comments

🍕 I recently wrote a High-Level Design (HLD) for a production-scale food delivery platform like Zomato / Swiggy 🚀 Food delivery looks simple on the surface 👀 — but under the hood it’s one of the most complex consumer systems to build: 🔥 millions of users placing orders concurrently 📡 real-time order & delivery tracking 💳 payments, refunds & dynamic pricing 🏪 restaurant & delivery partner coordination ⚡ high availability under peak load In this write-up, I cover: 🏗️ end-to-end system architecture 🧠 service boundaries (users, orders, payments, tracking) 🔁 real-time event flows 🚀 caching, messaging & data stores 📱 frontend + backend interactions 📄 Read it here 👉 🔗 https://lnkd.in/g3iNG8Qg If you’re preparing for system design interviews 🧑‍💻 or working on large-scale consumer platforms 🌍, this is exactly the kind of architecture you need to think about. 💬 Would love to hear your thoughts or discuss trade-offs 👇 #systemdesign #architecture #engineering #microservices #scalablesystems #fooddelivery

104

3 Comments

In my previous post I mentioned Jio was blocking my Supabase URL, and here's how I resolved it. Building in India sometimes means that your ISP decides your backend doesn't exist. Jio was silently blocking my Supabase domain, disrupting authentication, queries, and everything else for users on their network. The solution? A Cloudflare Worker as a proxy. This required about 30 lines of code, no changes to app logic, and just one environment variable update. The Worker intercepts all requests, rewrites the hostname to Supabase, and forwards them. Jio sees Cloudflare traffic, which it cannot block without affecting a large portion of the internet. Here are three challenges I encountered along the way: - Handling OPTIONS requests, as CORS preflight silently fails - Using redirect: 'follow' to ensure Supabase auth returns JSON instead of HTML - Ensuring there is no trailing slash on the Worker URL in your .env If you've faced ISP blocks on other services, this same pattern can be applied universally. #Supabase #StartupIndia #ProductLife #BuildInPublic #SaaS #EduTech

15

Everyone wants to become a founder. No one wants to be an intern. Many feel that working under someone is a low thing. But the truth is, no matter what you do, you are always answerable to someone. In a job, it is your manager. In a startup, it is your customer or investor. As a freelancer, it is your client. The real skill is not about escaping people. It is about learning how to work with them, grow, and deliver results. So stop ignoring roles just because they sound small. They may be the ones that actually teach you how to grow.

441

26 Comments

🔥 𝗙𝗿𝗼𝗻𝘁𝗲𝗻𝗱 𝗹𝗼𝗼𝗸𝘀 “𝘀𝗶𝗺𝗽𝗹𝗲”… 𝘂𝗻𝘁𝗶𝗹 𝘆𝗼𝘂’𝗿𝗲 𝘁𝗵𝗲 𝗼𝗻𝗲 𝗯𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝗶𝘁. People only see the UI, so every tiny issue shows up instantly — and that’s what makes frontend challenging. 🔁 𝗦𝘁𝗮𝘁𝗲 𝗺𝗮𝗻𝗮𝗴𝗲𝗺𝗲𝗻𝘁 A single incorrect update can trigger UI flickers, stale data, or race conditions that users notice immediately. 🧩 𝗥𝗲𝘂𝘀𝗮𝗯𝗹𝗲 & 𝗳𝘂𝘁𝘂𝗿𝗲-𝗽𝗿𝗼𝗼𝗳 𝗰𝗼𝗺𝗽𝗼𝗻𝗲𝗻𝘁𝘀 One poorly designed component becomes technical debt across teams, since everything depends on it. 📦 𝗗𝗲𝘀𝗶𝗴𝗻 𝘀𝘆𝘀𝘁𝗲𝗺 𝗰𝗼𝗻𝘀𝗶𝘀𝘁𝗲𝗻𝗰𝘆 Small spacing or alignment mismatches across teams, browsers, and devices quickly break the feel of the product. ⚡ 𝗣𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲 𝗯𝘂𝗱𝗴𝗲𝘁𝘀 Even one heavy image or unnecessary re-render can degrade performance and reduce conversions. 🎯 𝗣𝗶𝘅𝗲𝗹 𝗽𝗲𝗿𝗳𝗲𝗰𝘁𝗶𝗼𝗻 Layouts must work flawlessly across desktops, mobiles, tablets, foldables, and zoom levels without breaking design fidelity. 😩 𝗟𝗮𝘁𝗲 / 𝗶𝗻𝗰𝗼𝗺𝗽𝗹𝗲𝘁𝗲 𝗱𝗲𝘀𝗶𝗴𝗻 𝗵𝗮𝗻𝗱𝗼𝗳𝗳𝘀 Missing edge cases and last-minute tweaks often cause major rework and unexpected complexity. 🧪 𝗨𝗜 𝗯𝘂𝗴𝘀 𝗮𝗿𝗲 𝗲𝗮𝘀𝘆 𝘁𝗼 𝘀𝗽𝗼𝘁 A 3px movement, color mismatch, or misalignment immediately stands out and hurts the user's trust. 📉 𝗙𝗿𝗼𝗻𝘁𝗲𝗻𝗱 𝗲𝗿𝗿𝗼𝗿𝘀 𝗮𝗿𝗲 𝗽𝘂𝗯𝗹𝗶𝗰 Backend bugs hide in logs; frontend bugs show up on the user's screen, screenshots, and social media. 💡 𝗙𝗿𝗼𝗻𝘁𝗲𝗻𝗱 𝗶𝘀𝗻’𝘁 𝗲𝗮𝘀𝘆 — 𝗶𝘁'𝘀 𝗵𝗶𝗴𝗵-𝘃𝗶𝘀𝗶𝗯𝗶𝗹𝗶𝘁𝘆 𝗲𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝗶𝗻𝗴. Backend builds the logic — frontend builds the perception, and perception is everything. #frontend #frontendengineering #webdevelopment #javascript #reactjs #uiux #softwareengineering #webdev #designsystems #performanceengineering #developerexperience #tech

24

1 Comment

10 indian companies that you might not know were yc companies: 1. groww (W18) 2. meesho (S16) 3. razorpay (W15) 4. zepto (W21) 5. orange health labs (S20) 6. khatabook (S18) 7. cashfree payments (S17) 8. lokal (S19) 9. okcredit (S18) 10. innov8 (S16)

126

1 Comment