Activity
2K followers
-
Zaid M. shared thisWe're excited to host an in-person tech talk in our NYC office on Nov 21st! Please register for an evening of networking, food, drinks, and tech talks about GAI in Ad Tech from Terry Tong and Sean Wiser. Looking forward to seeing some familiar faces. https://lnkd.in/dac2sY4KZaid M. shared thisWe're excited to bring back tech talks to our NYC Office. Join us for an evening of talks, discussion, and networking. Learn how engineers at LinkedIn are leveraging GAI in ad tech. Food and drink will be provided at our LinkedIn NYC office at the Empire State Building.
-
Zaid M. shared thisADP’s next-generation platform powers dynamic teams with Amazon Neptune Check out our virtual AWS re:Invent session airing at 12:30pm ET today! Kuntal Roy and I are presenting. https://lnkd.in/gk2ZpmN ADP’s Next Gen Human Capital Management (HCM) platform embraces how work really gets done—in dynamic teams that do not reflect traditional organizational hierarchies. Running over 200 microservices on Amazon Elastic Kubernetes Service (Amazon EKS), the platform leverages Amazon Neptune to model and query its dynamic team structures. The session covers why ADP picked Amazon Neptune, followed by technical deep dives on how ADP bulk loads and synchronizes graph data, and what it learned using Amazon Neptune with Amazon EKS.
-
Zaid M. shared thisExciting times for us here at LifionZaid M. shared thisWhile we at Lifion stay focused in our core mission it’s important to take time out to celebrate these little achievements and acknowledgements from industry analysts around our work for the past 3-3.5 years. The engines have started,its time to take it up a notch.#Lifion
-
Zaid M. shared thishttps://lnkd.in/dS-rTRW Please message me if you're interested in applying!
-
Zaid M. shared thisWe'll be doing an information session at Waterloo on Jun 18! https://lnkd.in/eGGu8Ji
Zaid M. commented on a post
10mo
Hello!
-
Zaid M. liked thisZaid M. liked thisCareer Update: I'm heading back to Nelnet, rejoining the FACTS team as an Engineering Director. I'm looking forward to building again in the education space and taking on new challenges. It will be great to work alongside some familiar faces while getting to know new teammates. Thank you to everyone who reached out and offered support along the way. I appreciate all of the conversations and connections that helped me get here.
-
Zaid M. reacted on thisZaid M. reacted on thisMost people have no idea how to use genetic testing in a longevity practice. I include a lot of clinicians in that, and for years I included myself. Part of the trouble is the words. "Genetic testing" is really several different tests doing several different jobs, and we keep treating them as interchangeable. A consumer ancestry kit hands you raw recreational data. Pharmacogenomics tells you how to dose a drug. Single-gene tests flag rare, high-penetrance variants like APOE or BRCA. Polygenic risk scores give you a probability for something like heart disease. Clinical decision-support panels like IntellxxDNA read 600+ actionable variants and model how they interact, gene to gene and gene to environment. Run the wrong test and you get an answer to a question you never asked. Deepti Agarwal MD, DABOIM, MSCP and I are sitting down for a live conversation with Sharon H., co-founder of IntellxxDNA and a Harvard-trained physician. We'll discuss: -When to reach for genomics. -Which test does which job. -How to read the results without fooling yourself -More
-
Zaid M. liked thisZaid M. liked thisI've been writing Swift in production. I am not an iOS engineer. That's the thing about startups. The interesting problems don't stop at your job title. Some context. Harmonic is the startup discovery platform. Scout, our AI research agent, helps investors, companies, and talent find early-stage startups and founders to back, doing the research in the background while you work. We built Scout on the web. Bringing it to mobile sounded easy; a few Claude prompts to generate Swift files and a confident nod. Instead I found the hard part isn't Swift. It's figuring out what an AI agent should feel like when it lives across devices. By the way, you already do this. You search for a flight on your phone, half watching TV. Then you open your laptop to book it. One decision, two devices. You'd be annoyed if either one forgot where you were. Our users do that with companies. But agents add a twist: they don't just preserve state, they create it. Ask Scout to look into a company between meetings and it keeps digging, founders, customers, hiring signals, while you switch devices. So the challenge isn't only restoring where you left off, it's reconnecting you to work that continued while you were gone. The agent shouldn't feel tied to a tab or a screen. It should feel like one conversation across whatever interface you're on. That sounds obvious until you try to build it. A few things get weird fast: 1. Agent state isn't application state. Traditional apps sync user actions; agentic products sync ongoing work. What matters isn't where the user left off, it's where the agent is right now: a process that may have evolved while the app was gone. 2. Mobile systems weren't built for persistent agents. Agents want to run for minutes. iOS wants to suspend in seconds. So the agent can't live on the device. Continuity gets rebuilt from server-held state, background sync, push, and streaming. 3. Agent output is probabilistic. Mobile UX isn't. Native apps feel trustworthy because they're deterministic: tap a button, get a result. Agents don't work that way. You don't know what the final answer looks like, or how long it'll take. So what should stream? What should be optimistic? What should persist across devices? A native app has to feel immediate and reliable while the thing underneath it is neither. We're all still figuring out what these experiences should be, and getting to design both sides, the agent and the interface, is exactly why it's fun. (How we rebuilt Scout for this: link in comments.) The part where I tell you we're hiring: We're bringing on a Senior iOS engineer to own this end to end. Not just the architecture and patterns, but the bigger question: what agentic experiences should exist for our users, and how they translate to truly native iOS. Partly because the problem deserves someone who's spent a career getting good at this. Partly so someone qualified can review my Swift. If any of this made you lean forward slightly, my DMs are open.
-
Zaid M. liked thisZaid M. liked thisHosting the annual Ad Serving and LAN potluck was a wonderful experience. It brought together the team, cross-functional partners, and alumni in a relaxed setting. Connecting outside of work for a few hours led to engaging discussions, spontaneous brainstorming, and long-overdue catch-ups with some alumni. This gathering was definitely an energy refresh for me. #linkedinlife #dreambig #getshitdone #knowhowtohavefun #dreamteam #linlkedads Ayusman Sarangi Ilesh P. Shadi Rasheed Ada Yu Alexander Gavrilov Gaurav Sharma Ming (Mark) Yang Lin Du Jiayi Liu Yaolin Wang Rahul Srivastava Jill Wang Mariana Shargorodsky Xin Zheng Gayathri Gopalakrishnan Aakash Dhongade Christine Chou Rooney Jingyuan Gao Yanping Chen Jessica Yuan Xi Xiong Ruoying Wang Matthieu Hugues-Nuger Yanbo Ma Zhongwei Jiang Zihan (Crescent) Xiong Amanda Lee
-
Zaid M. liked thisZaid M. liked thisThrilled to share that I've joined Netflix! 🎉 A little on why this feels so personal to me — I grew up in Suwon, a small town about an hour from Seoul, watching K-dramas with my family. My love for a great story has been the one constant throughout my life. To this day, I'm always chasing the next great series, film, or book — and that passion has spilled into creating my own stories too. I have two completed novels and am hoping to self-publish one this year. My career has taken me through the ads world — seller ads at eBay, ads optimization at Meta, and ads measurement at LinkedIn — and Netflix sits at the exact intersection of two things I'm passionate about: entertaining the world with great stories, fueled by advertising. The mission clicked for me immediately. I couldn't be more excited to be here. Here's to the next chapter. 🍿
-
Zaid M. liked thisZaid M. liked thisWhile packing up my place in California for the move to Saudi Arabia, I found this old TP-Link router amongst my 12 years' worth of accumulated stuff. It brought back one of my favorite debugging memories from my time on the YouTube Android player team. Back in 2014 when we were rolling out ExoPlayer, we had a bug that the users were hitting in the wild, but we could not reproduce reliably in the office. My hunch was that it had something to do with real-world network conditions out in the wild: low bandwidth, packet loss, flaky connectivity. The kind of messy reality Google's amazing office connectivity usually protects you from. So I found a used TP-Link router on a corp for-sale group, flashed OpenWRT on it, and built a tiny 'bad network' lab. I routed it through the guest Wi-Fi to keep it off the internal corp network, and set up traffic shaping to constrain the bandwidth and drop packets at random. I named the Wi-Fi SSID to "$5/hr" as a little deterrence strategy. If you create a random-looking SSID, people get curious. If you create something that looks like a plausible paid hotspot, most people ignore it and move on. It made the network look boring enough that nobody would connect to it casually. Once I started testing against it, the bug reproduced fairly quickly. Clean environments hide dirty bugs. Sometimes the fastest way to understand your system is to recreate the messy reality your users are living in. Not just my bug, the setup exposed a few other issues in our ABR logic that were hard to see under clean network conditions. Soon, other engineers and QA folks started using it too. At one point, a QA specialist filed a bug with reproduction steps that literally started with: “1. Connect to the Wi-Fi network $5/hr.” That was the moment I realized my little debugging hack had become a little too official. Then one Friday, while I was away for Friday prayers, I got a message from a colleague: "Hey, the Wi-Fi police are at your desk asking whose router this is." I came back expecting an uncomfortable conversation. Instead, the person was very reasonable. He asked why I was running it, I explained the need, and he basically said, "We know teams need this kind of testing. We are working on a proper solution. But please take this one down." Fair enough. And to their credit, they did deliver a "Google Impaired Network" we could use for testing these kinds of scenarios properly. This was part of Google's engineering culture: when the right tool didn't exist yet, people found creative ways to get the work done.
-
Zaid M. liked thisZaid M. liked thisWe’re coming to the end of half, which means several projects are coming to an end. Creators on LinkedIn are hopefully going to like what they see! Stay tuned 😎
Other similar profiles
Explore more posts
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content