Ajay Chavan @TheOnce_

@akshad_999 I use below for state management - Zustand for dialogs, drawers, etc - React hook form with zod for form state management - React query for server side state management

x.com/UpSkillYourLif…

TypeScript dev watching JavaScript dev fixing “cannot read properties of undefined”

Big news… As of 1 month ago, I am officially working full-time on @Tan_Stack! It’s been amazing to see it grow from a tiny react-specific table library into the multi-framework brand that it is today, but I'm even more excited about the future. TanStack is just getting started. 🧵

After 3 months of development, the first alpha example of @tan_stack Start was deployed today! 🎉

My favourite thing from the React 19 release? useRef's typings got fixed. It now requires an argument, and won't become randomly readonly when you pass it null. 🥲 finally

A lot of folks have been asking me why the TS 5.5 release is so exciting. Let's thread up.

Typed routes in Next.js

UUID "Versions" make about as much sense as the Kingdom Hearts timeline

Over the past 3 years, I've watched over 10,000 videos on Youtube. And the truth is, 94% of them were a complete waste of time. But these 8 channels are worth more than a 4-year degree: 🧵

I’m obsessed with great leadership. But when I was young, I wasted years and opportunities not knowing what great leadership looked like. Learn from my mistakes. 21 clear signals you have a great leader:

ML System Design #3 How Does Uber Use Machine Learning to Solve Their ETR Problem? They use gradient-boosted trees and Queues! ETR(estimating the time to request) calculation is essential to make sure both drivers and riders get the best user experience under all conditions. When supply(drivers) is high, drivers will need to wait significantly more to get a ride, and when demand(riders) is high, riders will have to wait more. Uber needs to estimate ETR to make sure the experience is smooth. And they do it with the help of 3 models. Step 1: The Supply Model The journey begins with the supply model, which uses a gradient-boosted tree to estimate the last person's position in the FIFO queue. This model digests real-time data on queue length and driver behaviors, like the rate of trip abandonments and priority queue joinings, to gauge the queue's supply status. Step 2: The Demand Model Parallel to this, the demand model predicts the rate at which queued drivers will be matched with incoming trip requests. This is where the temporal and external variables shine—flight schedules, weather conditions, and app engagement rates are all factored in, using another gradient-boosted tree to forecast demand in finely sliced 15-minute intervals. Step 3: Consumption Algorithm The final step is a simulation algorithm that integrates outputs from the supply and demand models. It virtually progresses through the demand array against the supply figure, simulating the queue's consumption. This process crystallizes into a wait time classification: short, medium, or long. Important to note that this system handled ETR only for airport-bound drivers and riders since they hold a significant portion of Uber’s supply

If you're looking to build a SaaS, here's what I'm using: - @MongoDB (database) - @nuxt_js (code) - @stripe (subscriptions) - @render (servers) - @cannyHQ (feedback) - @crisp_im (support) - @transloadit (media processing) - @PlausibleHQ (analytics) - @getRewardful (affiliates) - @SendGrid (email API) - @dubdotco (link shortening) Almost all have a free plan to get you started.

Kubernetes Architecture is the key core concept one "must" be aware to crack a DevOps Role. Everything you need to know about K8 architecture as a DevOps engineer in 8 minutes✨ A thread ... 🧵 Day02/30 of #30daysofdevops

What people think Apache Kafka is: - a dumb pipe powered by simple log data structure 🤕 What Apache Kafka actually is: - a beast of a streaming platform consisting of 1.2 million lines of code 💪 I spent hours analyzing the codebase. Let's dive into the data 🧵

10 must read white papers if you are a Software Engineer: 1) The Google File System: lnkd.in/g6tVzTtn 2) Dynamo: Amazon’s Highly Available Key-value Store: lnkd.in/giV8etGM 3) Paxos: The Part-Time Parliament: lnkd.in/g3S5CnDe 4) MapReduce: Simplified Data Processing on Large Clusters: lnkd.in/g9PeMYnP 5) Kafka: a Distributed Messaging System for Log Processing: lnkd.in/g8e38kgk 6) Spanner: Google’s Globally-Distributed Database: lnkd.in/gNYrrxyT 7) Bigtable: A Distributed Storage System for Structured Data: lnkd.in/gdSnJ77z 8) ZooKeeper: Wait-free coordination for Internet-scale systems: lnkd.in/gzUpDFhd 9) The Log-Structured Merge-Tree (LSM-Tree): lnkd.in/gvnr9Cvp 10) The Chubby lock service for loosely-coupled distributed systems: lnkd.in/gU9WpA4C What else would you add to this list? *** If you found it valuable: ♻ Repost to help others in your network. 🔗 Bookmark it for future reference. 🔔Follow @ashishps_1 for more such content. Join my free newsletter to learn coding and system design visually: newsletter.ashishps.com

I failed so many coding interviews until I learned these 30 lessons: A Thread 🧵

I used to be stressed out of my mind… Now, I only work 4 hours a day and make $730,000/month. 20 sentences that helped me 10x my productivity:

Genuine question. Is there any reason to use axios in environments where fetch is available? Other than familiarity.