Activity
13K followers
-
Vineel Y. shared thisKnown stuff but certainly worth reminding: Fully diagnose the root cause before slapping on expensive and often unnecessary fixes.Vineel Y. shared thisSomeone asked me what “a knack for squashing hard bugs” means on my LinkedIn profile. A knack is an ability that is hard to explain or teach. I have a knack for troubleshooting. It is hard to pinpoint what gave rise to that ability, but I could point to a few early lessons that forged it. Today, let’s focus on an expensive lesson I learned early on. As a poor college kid, I rebuilt the engine of my old ‘84 Mitsubishi Cordia in the garage of my apartment in Oakland. The car was running poorly, with 100K miles on it. I was broke, so I took it upon myself to fix the problem. There was no YouTube back then. All I had was an official factory service manual. The rebuild took the whole summer. I did it the proper way, too, with full machining of the engine block and head, new piston rings, rebuilt carburetor, etc. It should have been a proud accomplishment. In a sense, it was, and this is what I tell people to this date - that I had rebuilt an engine. But that is not the whole story. After the rebuild, the car ran a bit better, but it still ran poorly. The bug laid elsewhere. An experienced mechanic down the street told me to check the catalytic converter. It was old and clogged, preventing good exhaust flow. After replacing the cat, it was like a different car. What did I learn? Fully diagnose the root cause before slapping on expensive and often unnecessary fixes. #systemdesign #debugging #troubleshooting
-
Vineel Y. shared thisVineel Y. shared thisI was a part of the fraud detection team at Amazon. Machine Learning algorithms were naturally a big part of the job. Most of us on the team were recent grads and extremely pumped up to use cool new tech to solve problems. There was once a specific classification problem that we were trying to solve that involved complex domain names. We spent three months building and training a model to do this, many days tweaking the parameters and finally ended up with a model that had an accuracy of ~70%. It was not great but not bad as far as models go. The problem was that the 30% false positives meant an operational overhead that we couldn’t sustain at our scale. Enter new hire Business Analyst. Day 1. He looks at our model, understands the problem that we’re trying to solve and then sits at his computer for a couple of hours and comes back and tells us it’s done. We’re not sure what he means by done so we go over to look and see something that looks like the image below. The guy basically built a set of regexes to solve the problem and it worked with a >99% accuracy. We felt like such idiots that day. Over the years I've seen hundreds of people, to their eventual peril, trying to force ML solutions to deterministic problems; the classic cases of solutions in search of problems. Cool new tech doesn't always make for the right solution. I learned a very important lesson that day; the smartest people are those who spend their time really understanding the problem, and then, come up with the most simple and elegant solution to it. Note: Screenshot is a stock image and NOT the real regex used.
-
Vineel Y. shared thisProp 19 Good work on posting this Sonia VuVineel Y. shared thisProp 19, How Does It Impact You? Watch this video for more info. . . 𝑾𝒉𝒐 𝒀𝒐𝒖 𝑾𝒐𝒓𝒌 𝑾𝒊𝒕𝒉 𝑴𝒂𝒕𝒕𝒆𝒓𝒔 ! 🌐 https://www.soniavu.com 📧 soniavurealestate@gmail.com Call / Text: 408.687.2987 #realtor #topproducer #justlisted #homesinsanjose #sanjosehomes #realestateinvesting
-
Vineel Y. shared thisVineel Y. shared thishttps://lnkd.in/ebnFYkm
-
Vineel Y. shared thisVineel Y. shared thisWe are proud to be one of The Silicon Review's Top 50 Growing Companies of the Year! Read more on Telenav's history and the road ahead:
-
Vineel Y. shared thisTelenav riding the automotive renaissance wave .Vineel Y. shared thisInvesting in Telenav is investing in the future of mobility. $TNAVTelenav Inc: Overlooked Tech Stock Up 292% in 2019, Has Strong OutlookTelenav Inc: Overlooked Tech Stock Up 292% in 2019, Has Strong Outlook
-
Vineel Y. shared thisVineel Y. shared thisAt Telenav, we believe in the future of mobility. AI-enabled automotive HMI software is predicted to reach $4.6 billion worldwide by 2025, according to a recent report by Tractica. See more on this #globalmarket growth in Yahoo Finance:
-
Vineel Y. shared this#TNAV upgraded to hold https://lnkd.in/g8TUADxTelenav (TNAV) Upgraded to Hold at Zacks Investment ResearchTelenav (TNAV) Upgraded to Hold at Zacks Investment Research
-
Vineel Y. reacted on thisVineel Y. reacted on this18,150 cuts in a single day. April 23. Meta cut 10%. Microsoft offered its first voluntary buyout in 51 years (rule: age + tenure >= 70). Nike cut 1,400, "majority in technology... using more advanced automation." I have 26 years in IT. Two restructuring rounds. Lesson: do not wait for the buyout email. Audit your own stack in June, not December. Babak Hodjat (Cognizant Chief AI Officer) said this week: "AI becomes the scapegoat... like when a company hired too many." Most cuts are ZIRP-era corrections. AI is political cover. But for a 45-year-old senior with 18 years tenure, the outcome is identical. They are not firing you for 26 years of CLI. They are firing you for zero hours with Claude Code. Pick one tool this week. Build something with it.
-
Vineel Y. liked thisVineel Y. liked thisI feel like there is a lesson here for those who use AI for their immigration case filings and also ask AI about the merit of their case. •••
-
Vineel Y. liked thisVineel Y. liked thisClaude Code launched in Feb'25 and look what it has done to software engineering jobs!
-
Vineel Y. liked thisVineel Y. liked thisIn large storage clusters, a single node itself may hold hundreds of terabytes. So scaling a cluster(either up or down) in storage is never a trivial operation as the nodes carry a large amount of persistent data. When a node is simply removed, the cluster immediately loses some replica redundancy and triggers a background replication to get back into safe state. It will take some time to get back to normal. That is why mature storage systems don’t “remove” nodes: they decommission them. Data is gradually and safely migrated to other nodes while maintaining fault tolerance. Growth brings a different challenge. As data footprints increase, operators typically have two cluster expansion choices: • Vertical scaling – add more disks to existing servers (if slots are available) • Horizontal scaling – add entirely new storage nodes Both are common in real deployments, and both introduce a new problem: data imbalance. After expansion, some nodes become hot while others remain underutilized. Without rebalancing, performance and reliability degrade over time. The good news is that Apache Ozone includes all of these tools. Post Scale Up: • Cluster balancer for horizontal expansion: https://lnkd.in/gDT6tm26 • Disk balancer for vertical expansion (adding new drives): https://lnkd.in/gCS7HAdm To Scale Down: • Safe node decommissioning for scale-down: https://lnkd.in/gSU_fYhR These features are not theoretical: they are operational tools intended for real production clusters. If you run Ozone (or are evaluating it), I encourage you to try these capabilities and share feedback with the Apache Ozone community at: https://lnkd.in/gkDhY6dF
-
Vineel Y. liked thisVineel Y. liked thisAI just saved me 3 hours on my US trip. No jokes! I was planning a trip to the Grand Canyon and the driving route made zero sense. So instead of accepting it, I called a few gpt agents and let them rethink the route from scratch. Result? A dramatically shorter path. Minutes instead of hours. Agentic AI for the win. Excited to share it with all of you! #ai #agents
-
Vineel Y. liked thisVineel Y. liked this🚨 New Blog Post Alert 🚨 Cluster-level balancing isn’t enough. Real-world ops create imbalance inside Datanodes—after disk replacement, expansion, or large deletes. Disk Balancer closes that last gap. A small step toward fully self-managing object storage. 👉 Read the preview (coming in Apache Ozone 2.2): https://lnkd.in/gmkprM-E #SelfHealingSystems #StorageEngineering #ApacheOzoneNo More Hotspots: Introducing the Automatic Disk Balancer in Apache Ozone | Apache OzoneNo More Hotspots: Introducing the Automatic Disk Balancer in Apache Ozone | Apache Ozone
-
Vineel Y. liked thisVineel Y. liked thisPost Graduate Program in AI and Machine Learning • Anupama Janakiram • SimplilearnPost Graduate Program in AI and Machine Learning • Anupama Janakiram • Simplilearn
-
Vineel Y. liked thisVineel Y. liked thisOver the past year, I’ve spent a considerable amount of time listening. To teachers and professors supporting every student in crowded classrooms. To principals balancing academics, daily operations, parent expectations, and what’s truly best for students. To parents worried about focus, real understanding, and whether learning has any deeper purpose for their children. And to child psychologists and psychiatrists who consistently spoke about rising concerns around attention, stress, and loss of curiosity in kids today. One thing became very clear. Not every child learns the same way, yet we continue to teach them the same way. So we started with a blank sheet of paper. No assumptions. Just honest questions about how learning should actually work for a child today. That journey led to SlateBook, a Web (slatebook.ai), iOS, and Android platform built to make learning more personal, meaningful, and supportive of each child’s mindset. Our belief is simple. Learning should build understanding, confidence, and clarity, not just complete syllabi. This is just the beginning, and I’m proud of what this team has built. I’m also deeply thankful to every educator, parent, and expert who took the time to share their perspectives and shape our thinking. If this resonates, I’d genuinely love to hear your thoughts. Learn more: https://slatebook.ai
Experience
-
-
-
-
-
-
-
-
-
-
-
Education
-
Birla Institute of Technology and Science
-
Course work in Computer Networks , Operating Systems, Data Structures, Algorithms and Software Engineering
-
-
Other similar profiles
Explore more posts
-
Redis
In-memory databases are the architectural backbone of real-time AI. They’re a necessary advantage for memory-intensive workloads that rely on instant access to context. Modern systems like Redis blend durability, multi-model flexibility (JSON, vectors, time-series), and tiered storage to deliver the speed of memory with the reliability of disk. The result? Real-time analytics, instant personalization, and AI systems that actually feel intelligent. Here’s how we deliver the speed of memory with the reliability of disk: https://lnkd.in/gSm65kdB
-
Bassirah Technology
AI systems struggle with accuracy and memory. Redis solves this with semantic caching, vector search, and scalable memory enabling agentic RAG at enterprise scale. As the official Redis distributor in Egypt, Bassirah is proud to help enterprises bring these advanced AI capabilities into production. #AI #DataManagement #DataIntegration #AIsolutions #TechSolutions
-
Prompt Leadz
Tip: Scale large HNSW vector sets in Redis by splitting your dataset across multiple instances and querying them in parallel with multiplexing, enabling efficient parallel reads and writes that drastically improve performance and manage memory effectively. This approach leverages Redis's scalable vector support to handle hundreds of millions of vectors seamlessly. #Redis #HNSW #VectorSearch #ScalableAI #DataScience
-
Networking Pulse
Redis Releases Critical Patch (CVE-2025-49844) A 13-year-old use-after-free flaw in Redis could allow remote code execution via crafted Lua scripts—affecting thousands of cloud instances. 🔹 Redis powers ~75% of cloud environments 🔹 Patch immediately & review Lua configurations 🔹 Audit access and monitor for suspicious activity NetworkingPulse can assist with patch validation, hardening, and threat monitoring. https://lnkd.in/efDZTvWC #Redis #Security #NetworkingPulse
-
LearnKube
Master Kubernetes with LearnKube's Advanced Kubernetes workshop! What should you expect? - Learn how to architect and design clusters from the ground up (in the cloud or on-prem). - Explore the Kubernetes internal component and how the system is designed with resiliency in mind. - Deep-dive into the networking components and observe the packets flowing into the cluster. - Hands-on labs to test the theory with real-world scenarios! - And more. The next course starts in September: https://lnkd.in/ge4aYVZq We also run in-person courses and private training: https://lnkd.in/g9N3m-Xr
-
Forem
The Redis AI Challenge ends in less than a week (Aug 10th deadline)! Think it's too late to join? Think again! Here are 3 project ideas that you can build in a week. Project 1: Real-Time Poll App Prompt: Beyond the Cache Build a web app where users can create polls and vote, with results updating live for everyone. A perfect project to showcase Redis for real-time applications. Redis Use: Hashes for poll data, Sorted Sets for vote counts, and Pub/Sub for live updates. Project 2: Simple Job Queue Prompt: Beyond the Cache Create a background job system. A web app submits a task (like sending an email) to a queue, and a separate worker process handles it. This is a core backend development pattern. Redis Use: Redis Lists as a simple and reliable message queue using LPUSH and RPOP. Project 3: AI-Powered Article Suggester Prompt: Real-Time AI Innovators Build a blog that recommends similar articles to the one being read. It's a fantastic, beginner-friendly entry into the world of AI. Redis Use: Use Redis as a vector database to store AI-generated embeddings of your articles and perform lightning-fast similarity searches. --- Head to the challenge page to see entry details and peep what other people are building. Can't wait to see what you build! https://lnkd.in/eH_Dk2wu #Redis #AI #VectorDatabase #MachineLearning #Backend #Programming
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content