About
Need a referral at Razorpay? If you utilise below referral link, it will count as my…
Activity
10K followers
-
Gagandeep Singh Ahuja reposted thisGagandeep Singh Ahuja reposted thisFrom a Naxal-affected village in Chhattisgarh to becoming an IAS officer, Namrata Jain's journey is a testament to resilience and purpose. After securing AIR 12 in UPSC, she chose to serve where help was needed most. By building 24 schools and bringing education, healthcare, and essential services to remote communities, she has transformed thousands of lives. Her story proves that determination can overcome even the toughest beginnings. #InspiringIndia #PositiveNews #WomenInLeadership #UPSCSuccess #ChangeMakers [IAS Officer Success Story, Namrata Jain IAS, Women Empowerment India, Rural Education Development, Inspirational UPSC Journey]
-
Gagandeep Singh Ahuja reposted thisGagandeep Singh Ahuja reposted thisWhen tourists visiting Rajasthan’s Jawai region spoke about the growing plastic waste polluting its landscapes, Kana Ram Mewara couldn’t ignore it. So the tea seller from Bisalpur started collecting plastic himself and soon came up with a simple idea: people could exchange plastic waste for essentials like sugar, stationery, and groceries. Despite criticism and limited resources, he kept going. What began at a small chai stall has now grown into a 15-village movement that has helped collect over 15,000 kilograms of plastic waste and transform it into benches, dustbins, and tree guards. Loved the video? Stay tuned for more in #ForceForGoodHeroes, an extraordinary series on India’s unsung heroes that will inspire you to believe in the power of change. #forceforgood #adityabirlagroup In partnership with Aditya Birla Group
-
Gagandeep Singh Ahuja reposted thisWhy should the Cockroach Janta party and Sonam Wangchuk not demand his immediate release along with Dharmendra Pradhan's resignation? Is this how you treat a teacher, simply because he is a Muslim? And then you run for your life when a foreign journalist poses a question to you about your rock bottom position as a democracy. No shame? Not even a bit? So lost are you on shooting silly melody selfies.Gagandeep Singh Ahuja reposted thisModern India He became traitor for questioning to those in the power and their media
-
Gagandeep Singh Ahuja reposted thisGagandeep Singh Ahuja reposted thisFollowing the NEET paper leak controversy, Alakh Pandey strongly criticised the examination system and questioned how such a large-scale leak could happen despite strict rules imposed on students. In a latest conversation with News Pinch's Founder Abhinav Pandey, the Physics Wallah founder said students were being excessively monitored before exams from dress codes to transparent water bottles while people inside the system allegedly enabled the paper leak itself. “It’s you versus the system. The system was compromised,” he said, arguing that the issue reflected a deeper institutional failure rather than isolated malpractice. Pandey questioned why authorities did not change the paper even after initial indications of a possible leak and criticised the handling of the situation by the National Testing Agency. He also demanded strict punishment for everyone involved, including wealthy families allegedly purchasing leaked papers and any coaching institutes found complicit. “This is a national-level leak. People inside the system are involved,” he said, adding that the arrival of the CBI investigation could expose larger irregularities. The educator further highlighted the emotional and financial impact on students, saying years of preparation had been affected while taxpayers would ultimately bear the cost of conducting examinations again. #AlakhPandey #Neet #ReExamination #NTA #Insiders #Examination #Startupro
-
Gagandeep Singh Ahuja reposted thisGagandeep Singh Ahuja reposted thisICICI Lombard is denying my kidney transplant. I have 3 days to live. I am Om Prakash. I have end-stage kidney failure (eGFR 6). My mother, Mrs. Manti Devi, is my donor. My admission for transplant is 13 May 2026 at Medanta Hospital. My insurance policy: ICICI Lombard Elevate – Policy No. 100005295600 – has a sum insured of ₹10,00,000. Donor expenses are explicitly covered. But ICICI Lombard denied my cashless approval – twice. Their reason? "Ailment history not clear." Here is what my doctor wrote (13 April 2026): "Chronic Kidney Disease – recently detected on 09/04/2026 – NO past history of any disease or symptoms. " Here is what ICICI Lombard's own policy says: Pre-existing disease clause only applies if diagnosed in the 36 months BEFORE policy start. My policy started 16 May 2025. My diagnosis was 11 months AFTER that. No waiting period applies. Basic Cover 8 – Donor Expenses: Explicitly covers "organ donor's hospitalization for harvesting of the organ donated to the Insured Person up to the Annual Sum Insured." My mother is covered. This is not a gray area. This is clear. Yet, ICICI Lombard closed my grievance (Ref. 9270420264842) without resolution. ICICI Lombard, Sanjeev Mantri – I am asking you directly: Will you wait for me to die on the operating table before you approve my cashless? Insurance Regulatory and Development Authority of India – Please investigate this violation of cashless guidelines. I have attached my doctor's letter and the policy screenshots. My life depends on your response. 3 days left. ICICI Lombard ICICI Bank ICICI Prudential Life Insurance Company Limited Indian Ministry of Finance Department of Consumer Affairs #ICICILombard #HealthInsurance #KidneyTransplant #InsuranceDenial #IRDAI #PatientRights #DonorExpenses
-
Gagandeep Singh Ahuja reposted thisGagandeep Singh Ahuja reposted thisFraud behind NEET, UPSC, SSC Exams in India (and the rising concerns) Naman Shrivastava
-
Gagandeep Singh Ahuja reposted thisGagandeep Singh Ahuja reposted thisI stood at the border so you could sleep peacefully. Today, at 74, I cannot sleep at all. My name is Satnam Singh. 32 years in the Indian Army. Monthly pension: ₹53,000. I have lost ₹1.28 crore to organised insurance fraud. I am ₹50 lakh in debt. What happened: In 2020, fraudsters impersonating an SDM and a senior IAS officer pushed me into fake IPO schemes and phoney pension plans. Policy after policy in my name, my family's, even my neighbours' all on forged documents. What the documents show: ▶ My income falsely inflated to ₹6–8 lakh. One policy's annual premium alone was ₹6.30 lakh my entire year's pension. ▶ My 18-year-old granddaughter listed as a farmer earning ₹6 lakh/year. ▶ My housewife daughter-in-law shown earning ₹12 lakh/year. ECS activated without her consent. ▶ My nephew's premium equalled 105% of his actual income. Still approved. The question nobody asked: I was 74. Why were long-term, high-premium policies issued to me at all? No suitability check. No income verification. Just commissions. Three companies are still silent holding ₹44.25 lakh of my money: ❌ Bharti AXA Life Insurance ₹22,70,000 ❌ Axis Max Life Insurance ₹11,25,000 ❌ Edelweiss Life Insurance — ₹10,30,000 ❌ Bharti AXA Life Insurance ₹22,70,000 Parag Raja, MD & CEO ❌ Axis Max Life Insurance Limited ₹11,25,000 Sumit Madan, MD & CEO ❌ Edelweiss Life Insurance ₹10,30,000 Sumit Rai, MD & CEO Their response? "The free-look period is over." Forged documents don't expire. IRDAI's KYC, AML, and policyholder protection norms were violated at every step. I am asking for what the law already guarantees: ✅ Cancellation of all fraudulently issued policies ✅ Full refund with interest ✅ Action against named agents ✅ Reporting to IRDAI and FIU-IND I never asked this country for anything in 32 years. Today, I ask for one thing — justice. Journalists, lawyers, consumer rights advocates & IRDAI officials — please reach out. 📞 Satnam Singh | Retired Army Officer | Punjab
-
Gagandeep Singh Ahuja reposted thisGagandeep Singh Ahuja reposted thisRehan: Yesterday was one of the worst nights in Indian history. When the PM of the country came to deliver what was supposed to be an address to the nation, it turned into a political speech. Whenever any head of the country announces that he is going to address the nation, the whole country stops and waits. People become tense, thinking it might be about war, an emergency situation, a major decision, or a big national project. It is expected to be a serious national message, not a political speech. On the contrary, what PM Modi did? He mainly criticised the opposition over a bill that had already been passed under his own government with a full majority the women’s reservation bill. In whole speech he was criticising opposition, bunch of lies and emotional drama. The whole nation has been misled and Fooled. If this continues, people may stop taking such national addresses seriously in the future. The Indian media does not seem to have the courage to question this, and neither does the ECI. It feels like Indian democracy is being controlled by one man and one party in a shameless manner.....almost like a banana republic.
-
Gagandeep Singh Ahuja reposted thisGagandeep Singh Ahuja reposted thisTwo years ago, a tanker carrying 20,000 liters of milk overturned near a river in the UK. Within minutes, 𝗲𝗺𝗲𝗿𝗴𝗲𝗻𝗰𝘆 𝘀𝗲𝗿𝘃𝗶𝗰𝗲𝘀, police, fire crews, paramedics, were all at the scene. They blocked the road. Built barriers. Made sure not a single drop reached the water. 𝗧𝗼 𝘁𝗵𝗲𝗺, 𝘁𝗵𝗶𝘀 𝘄𝗮𝘀 𝗮 𝗰𝗿𝗶𝘀𝗶𝘀. Now look at what we are doing in India, where most of our rivers and lakes are already dying. We are voluntarily pouring 𝗺𝗶𝗹𝗸 𝗮𝗻𝗱 𝗴𝗵𝗲𝗲 into them. This is when we know that ghee doesn't dissolve in water. It spreads, a 𝘁𝗵𝗶𝗰𝗸 𝗼𝗶𝗹𝘆 𝗹𝗮𝘆𝗲𝗿 across the surface, 𝘀𝘂𝗳𝗳𝗼𝗰𝗮𝘁𝗶𝗻𝗴 𝗲𝘃𝗲𝗿𝘆𝘁𝗵𝗶𝗻𝗴 beneath it. We are blessing a river we are slowly killing. ------------------- This is a reminder of why the 𝗿𝗶𝗴𝗵𝘁 𝘂𝗻𝗱𝗲𝗿𝘀𝘁𝗮𝗻𝗱𝗶𝗻𝗴 𝗼𝗳 𝗿𝗲𝗹𝗶𝗴𝗶𝗼𝗻 must reach people urgently. As 𝗔𝗰𝗵𝗮𝗿𝘆𝗮 𝗣𝗿𝗮𝘀𝗵𝗮𝗻𝘁 points out, real religion concerns itself with only one thing. You, and your suffering. Everything else is noise. And until we understand what religion is actually for, our rivers will keep drowning in pollution. And in piety.
Gagandeep Singh Ahuja commented on a post
1w
Congratulations Aditya! 💙 🎉
-
Gagandeep Singh Ahuja liked thisGagandeep Singh Ahuja liked thisSome chapters leave a lasting impact, not just because of the work, but because of the people you meet along the way. As I look back on my 1.5 - year journey at Razorpay , I'm grateful for everything it brought me. During this time, I had the opportunity to work across the POS BU (Omni) and the R1 BU, contributing to the Partnerships, Onboarding, and Rize teams. Each team came with its own challenges, and every transition pushed me to learn, adapt, and grow as an engineer. Product: To Pratul Gupta - thank you for the collaboration, thoughtful discussions, and for always keeping customer impact at the heart of what we built. Working together made every milestone even more rewarding. Managers: A special thank you to Apurva Gupta , Shobhit Jain and Ketan Somani for believing in me, trusting me with ownership, and creating an environment where I could continuously learn and take on new challenges. Your mentorship has had a lasting impact on my professional growth. Leadership: To Rizwanul Haque and Ashish Tewari - thank you for setting a clear vision and building a culture that encouraged innovation, ownership, and continuous learning. Engineering: To Shwetabh Shekhar , Sunny Aggrawal , Bhavya Y , Danish Iqbal , Manikanta Kandagatla , Venkata Subbaiah Machunuru , Sujan Adiga , Ketan Suryawanshi , Sahaj Dave , Shwatang Agarwal , Sathwik Kuriseti , Lokesh B M , Anup Kanere , Jeevith R , Ayushi Dubey , Abhinav Kumar , Srijan Baranwal , Arun kumar G E , Sai Phanidhar Ravinuthala , Ananya Agrawal - thank you for every code review, architecture discussion, debugging session, production incident, sprint, and release. The late-night fixes, shared victories, and everyday collaboration made this journey memorable. I learned something from each one of you, and I'm grateful to have worked alongside such an incredible team. Razorpay wasn't just a place where I wrote code - it was where I learned to solve problems at scale, work with exceptional teammates, and appreciate the value of strong collaboration. As I begin the next chapter, I'm taking forward not just technical learnings, but the relationships and experiences that made this journey special. Thank you, Razorpay. On to what's next. 🚀 #Razorpay #SoftwareEngineering #EngineeringCulture #CareerGrowth #Learning #Teamwork #TechLife #NextChapter
-
Gagandeep Singh Ahuja reacted on thisGagandeep Singh Ahuja reacted on thisThe last and final video of my Redis Internals series is out, and it answers one simple question: "Which keys should be evicted when memory is full?" The obvious answer is LRU or LFU. We looked at LRU some time back, so now let's dig into LFU. To implement LFU, the standard solution is to maintain an integer counter for every single key and increment it on every access. This approach is perfectly accurate, but it consumes a good chunk of memory (a 4-byte overhead for every single key). So how does Redis keep track of Least Frequently Used (LFU) items while using just a few bits of memory? That is exactly where approximate counting kicks in. Today, we dive into the internal mechanics of how Redis implements its LFU eviction strategy using the Morris Counter algorithm. This is the 26th and final video in the Redis Internals series. In the video, I talk about the core intuition behind LFU in Redis and how it packs both the last decrement time and a logarithmic counter into just 24 bits. I break down the math behind the Morris Counter, explain why we can afford to lose precision for higher frequencies, and look at how Redis handles time-based exponential decay without running a heavy background thread. By the way, all 26 videos of the series are out now: 1. Why Single-Threaded Redis Is Fast 2. Writing a TCP Echo Server 3. Wire Protocols 4. Implementing RESP 5. Implementing PING 6. Understanding Event Loops 7. Implementing Event Loops 8. Implementing GET, SET, and TTL 9. Implementing DEL, EXPIRE, and Cleanup 10. Evictions and Implementing first-eviction 11. Implementing Command Pipelining 12. Implementing AOF Persistence 13. Objects, Encodings, and Implementing INCR 14. Implementing INFO and allkeys-random Eviction 15. The Approximated LRU Algorithm 16. Implementing the Approx LRU Algorithm 17. How Redis Caps Its Memory Usage 18. How and Why Redis Overrides Malloc 19. Graceful Shutdown using Signal Handling 20. Redis Transactions Internals 21. Redis Memory-Efficient Lists 22. Redis Memory-Efficient Sets (Intset) 23. How Redis powers Geospatial Queries 24. How Redis Implements Strings 25. HyperLogLog and Cardinality Estimation 26. How Redis implements LFU using Morris Counter Thank you for following along with the series. I hope this helped you better understand database internals and sparked that engineering curiosity. Give it a watch.
-
Gagandeep Singh Ahuja liked thisGagandeep Singh Ahuja liked this𝗔𝗜 𝗜𝗻𝘁𝗲𝗿𝘃𝗶𝗲𝘄 𝗦𝗲𝗿𝗶𝗲𝘀 #16: 𝗖𝗵𝗮𝘁 𝗖𝗼𝗺𝗽𝗹𝗲𝘁𝗶𝗼𝗻 𝗔𝗣𝗜𝘀 — 𝗧𝗵𝗲 𝗔𝗯𝘀𝘁𝗿𝗮𝗰𝘁𝗶𝗼𝗻 𝗧𝗵𝗮𝘁 𝗠𝗮𝗱𝗲 𝗟𝗟𝗠𝘀 𝗨𝘀𝗮𝗯𝗹𝗲 One of the most common GenAI interview questions is: 👉 "What is a Chat Completion API?" Most candidates answer: > "It's an API that sends a prompt to an LLM and gets a response." Technically correct. But that explanation misses one of the most important architectural shifts in modern AI. ### Before Chat Completion APIs Traditional NLP systems worked like this: Input → Model → Output Example: ``` Translate this sentence. ``` Response: ``` Translated text. ``` Simple. Stateless. Limited. --- ### Modern LLM Applications Applications such as ChatGPT, Claude, and Gemini require much more: • System Instructions • Conversation History • Tool Calls • Function Execution • Structured Outputs • Context Management This led to the emergence of the Chat Completion API abstraction. Instead of sending a single prompt, we send a conversation. Example: ``` [ { "role": "system", "content": "You are a senior solution architect." }, { "role": "user", "content": "Explain Kubernetes." } ] ``` The model now understands: ✅ Who it should behave as ✅ What the user wants ✅ Previous conversation context ✅ Available tools and functions --- ### Why This Changed Everything The breakthrough wasn't building larger models. It was standardizing how applications interact with those models. Chat Completion APIs enabled: • AI Assistants • AI Agents • RAG Systems • Coding Assistants • Enterprise Copilots • Multi-Agent Workflows Without this abstraction layer, building production-grade LLM applications would be significantly more complex. --- ### Architect Perspective When designing enterprise GenAI systems, the model is often the easiest component. The real challenges are: • Prompt orchestration • Context management • Memory handling • Tool integration • Cost optimization • Conversation state management The Chat Completion API became the contract that allowed these systems to scale. --- ### Interview Takeaway Don't say: > "A Chat Completion API generates responses." Say: > "A Chat Completion API is a conversational abstraction layer over LLMs that manages roles, context, memory, and tool interactions, enabling the development of stateful AI applications and agentic systems." Modern AI isn't just about models. It's about the interfaces that make those models usable. #AI #GenerativeAI #LLM #AIAgents #ChatGPT #AIEngineering #SolutionArchitecture #RAG #MachineLearning #MLOps
-
Gagandeep Singh Ahuja liked thisGagandeep Singh Ahuja liked thisWorld-renowned therapist & NYT bestselling author Dr. Shefali explains how fathers can teach their daughters to handle the patriarchy. What are your thoughts on this? Let me know in the comments 💭
-
Gagandeep Singh Ahuja liked thisGagandeep Singh Ahuja liked this🚨 A SENIOR ANTHROPIC ENGINEER JUST DROPPED AN 11-PAGE PDF ON LOOP ENGINEERING. The core shift: stop prompting the agent. Build the system that prompts it. Inside the autonomous loop: - Discover → Finds its own work (failing CI, open issues). - Isolate → Uses separate git worktrees to prevent collisions. - Verify → A second agent reviews the work. (Never let agents self-grade). - Persist → Writes to disk, not temporary context windows. - Schedule → Runs automatically on a timer. This is a great framework for building more reliable agentic systems Link to the free 11-page guide → https://lnkd.in/e8BMrh8W Read it, then check out the ace article on Loop Engineering by Akshay Pachaar → https://lnkd.in/eGJBZEGw
-
Gagandeep Singh Ahuja liked thisGagandeep Singh Ahuja liked this𝗙𝗼𝗿 𝗱𝗲𝗰𝗮𝗱𝗲𝘀, 𝘁𝗵𝗲 𝗳𝗼𝗼𝗱 𝗶𝗻𝗱𝘂𝘀𝘁𝗿𝘆 𝗵𝗮𝘀 𝗵𝗮𝗱 𝗼𝗻𝗲 𝘂𝗻𝗳𝗮𝗶𝗿 𝗮𝗱𝘃𝗮𝗻𝘁𝗮𝗴𝗲. They know exactly what’s in your food. You don’t. They have food scientists, R&D labs, and marketing teams whose entire job is to make a product look healthier than it is.🫠 You have eight seconds in a busy aisle and a label written to confuse you. That asymmetry - that’s the whole game. And today, we’re ending it.✨ 𝗧𝗵𝗶𝘀 𝗶𝘀 𝗪𝗵𝗮𝘁𝘀𝗜𝗻𝗜𝘁. 🔍 Search any packaged product by name - or browse what’s already on WhatsInIt - and in seconds you’ll know: → What’s really inside - every ingredient, in plain English → Whether the claims on the front are true or just theatre → What the label is quietly hoping you won’t notice → Whether something cleaner exists on the same shelf No paid rankings. No paid placements on SafeShelf. The first time the information has been on your side of the table. Here’s what we believe 👇 Eating well shouldn’t require a chemistry degree. Trust shouldn’t be something brands can buy. And “what’s in my food?” should have an answer that takes seconds - not a Google rabbit hole at 11pm. We didn’t build a label scanner. We built a power shift. 𝗪𝗵𝗮𝘁𝘀𝗜𝗻𝗜𝘁 𝗶𝘀 𝗹𝗶𝘃𝗲 𝘁𝗼𝗱𝗮𝘆. 👇 🔗 whatsinit.in Watch the video. Then go check one product you eat every week. I promise you’ll never look at the shelf the same way again. I’m in the comments all day - tell me what you find.🤗
-
Gagandeep Singh Ahuja reacted on thisGagandeep Singh Ahuja reacted on thisLooking back, three shifts in how I think and learn had an outsized impact. I wish I had made them on day one. 1. became more data-driven 2. started seeking contrary opinions 3. started learning tougher topics Start doing this as early as you can; it directly impacts your career growth, acceleration, output, and outcomes. The sooner the better. Hope this helps.
-
Gagandeep Singh Ahuja liked thisGagandeep Singh Ahuja liked thisPostgreSQL (and other relational databases) needs to estimate how many rows a query will return because it directly impacts which execution plan it chooses. Pick the wrong plan, and your query could be thousands of times slower. Here's how the query planner actually works under the hood to estimate the number of rows... When you run a range query like "SELECT * FROM table WHERE column < 1000", PostgreSQL doesn't just guess. In fact, for numerical values, it maintains histogram buckets that divide your data into equal frequency ranges. The planner figures out which bucket your value falls into, calculates what fraction of that bucket matches your condition, and estimates the rows. For equality conditions like "column = 'CRAAAA'", the approach is different. PostgreSQL maintains a list of the most common values (MCVs) with their exact frequencies. If your value is in that list, it uses the stored frequency directly. If not, it assumes the remaining values are evenly distributed among all the non-common distinct values. For joins, the planner looks at statistics from both tables. It considers null fractions, distinct value counts, and whether values appear in MCV lists. The approach combines these factors to estimate the number of matching rows the join will produce. There is a bunch of simple math involved in this estimation, and PostgreSQL's documentation covers it pretty nicely. Have added it below. Hope you find it interesting.
Experience & Education
-
Razorpay
View Gagandeep’s full experience
See their title, tenure and more.
or
Licenses & Certifications
Recommendations received
4 people have recommended Gagandeep
Join now to viewOther similar profiles
Explore more posts
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore More