About
Activity
3K followers
-
Nico Lopez shared this
-
Nico Lopez shared thisJust wrapped up hosting an AI Tools Workshop at Meta Connect 🎉 Here are a few key takeaways that stood out to me: ✨ AI tooling is evolving fast. Most developers are already adding it into their workflows. The demand is undeniable. ✨ Product strategy needs a rethink. As the way developers interact with tools shifts, UX investments should reevaluate if traditional UI fits in the equation. ✨ AI education is still a big gap. Knowing about AI isn’t enough - hands-on adoption is the only way to discover how it truly fits into your workflow.This was true for me too. And yes, setting up ~40 computers for live demos was pure, beautiful chaos 😅, but watching developers hit those “wow” moments in VR made it worth it. Excited to see where developers take the tools we announced today!
-
Nico Lopez shared thisCome join our Meta Developer Support Booth at #Devcom & #Gamescom We will be hosting a Unity / Quest workshop and I'll be here to answer any Profiling / Performance Optimization question you might have! Particularly, we will be introducing our new Quest Runtime Optimizer tool my team just launched. Articles: https://lnkd.in/gVFg5tUu Tutorial: https://lnkd.in/gSarsstY Hope to see you there and feel free to pop by and say hi!
-
Nico Lopez shared thisAI in Optics is a fairly overlooked field. It has high promise in military and industrial applications. Our paper being presented at the "Emerging Topics in Artificial Intelligence" conference about an application of deep learning to wavefront correction. The link is posted in the first comment. SPIE, the international society for optics and photonics conference Chair Giovanni Volpe tweeted: #spieopticsphotonics #ai #optics
-
Nico Lopez shared thisYesterday was my last day as a Software Engineering intern at Facebook. I can't help but recall, how every year until my junior year I never applied to Facebook because I thought I wasn't "good enough". I told myself that I didn't come from Harvard or MIT so they wouldn't even be bothered to look at my resume. This was definitely not the case. I successfully finished my internship and won an internal Facebook hackathon because I bet on myself. Something I sometimes really struggle doing and want to become better at! I really can't thank Shaahid Shaik, Max Radermacher, Denis Orlikhin, Cesar Valdez, Carin Gan, Hillary Stangel, Carin Gan, and the rest of my team for making me feel part of Facebook #latinxintech #facebook #intern2020
-
Nico Lopez shared thisI’m beyond excited to share that I have accepted a Software Engineering internship position @Facebook in Menlo Park! I can't wait to spend this summer learning from some truly amazing people! Thank you, Emilio Arias, Kaitlin Wilson, and Andy Pan for all of your help in the process! I am really looking forward to starting my journey at Facebook! #facebook #internship #latinxintech
-
Nico Lopez shared thisA week ago, I had the opportunity to connect with leading Latinx engineers at the Society Hispanic Professional Engineers' National Convention. Being surrounded by passionate Latinxs in STEM is beyond empowering! I've been fortunate enough to have amazing mentors and opportunities that showcase that Latinxs do belong in STEM and that being an engineer is a realistic dream. This is not the case for all Latinxs, but I am hopeful for what the future holds! At SHPE, I witnessed so many companies taking affirmative steps in the right direction! Onwards! #latinxintech #shpefamilia #shpe19
-
Nico Lopez shared thisAfter 36 hours of relentless work, I am incredibly honored to have earned the 'Best Hardware Hack' at HackHarvard! Our team built a dog training robot (Train-O) that visually classified a dog's actions and reinforced desired behaviors by dispensing treats and pairing them up with the owner's voice. This was an extremely enjoyable and challenging team effort that put our hardware and software engineering skills at test! I wanted to thank IWU for sponsoring ACM and everyone at HackHarvard for organizing this event! #hackharvard #openCV #hackathon #teamwork
-
Nico Lopez shared this#GoodNewsOfTheDay The Online Encyclopedia of Integer Sequences (OEIS) just published the sequence I found in my Algorithmic Number Theory research project! You can read more about the "Quasi-Fortunate Numbers" in the following link: https://oeis.org/A305099
-
Nico Lopez reacted on thisNico Lopez reacted on thisAll this happened, more or less. Seven years ago, I joined Meta as a software engineering intern. I came back as an engineer, eventually made the leap into product through RPM, and left as a Senior PM. A lot happened in between. I had my first SEVs. I worked on products that were sunset. I watched a team I loved get dissolved, moved on, and eventually found my way back to the same space. And I learned something surprisingly freeing: failure is survivable. Some of the most important feedback I got early in my PM career came from Xinghao Cheng and You Zhou: You have great ideas, but you need to defend them. A PM has to own the direction of the team. Later, Samir Gupte and Abhinav Dubey pushed me to do the same thing at a much larger scope. At every stage, there were people who saw what I could become a little before I did. They gave me room to grow into it before I fully believed it myself. Over time, my understanding of leadership also changed. At first, I thought being a good PM meant running a great meeting. Then it meant leading a team. Then multiple teams. Eventually, it meant thinking about how organizations move. I used to think of a PM as the conductor of an orchestra: you don’t play every instrument, but you make sure everyone is playing together. Now I think the job starts much earlier. What music should we play? What orchestra do we need to play it? How do you give people a vision and enough of the sheet music to move together, while leaving room for them to make it better? How do you listen when they hear something you don’t, and ultimately conduct something none of you could have made alone? Meta taught me a lot about building products. But more than anything, it taught me judgment, clarity, and trust: how to form a point of view when the answer isn’t obvious, make a case for it, bring people along, and then give a great team the space to build. And somewhere along the way, something I’d known about myself for a long time became clearer: the part I love most is the very beginning. So after seven years, I’m leaving Meta to join a startup — with fewer guardrails, more uncertainty, and, yes, a little fear. But maybe that’s the point. Part of growing up is leaving home. Thank you to the mentors who challenged me, the teammates who trusted me, and the friends who made the seven years go by impossibly fast. I’m especially grateful to Akshay Mittal Divya Bhattarai Rob Veres Evelyn Lee Cole Vertikoff Yang Tang Soham Khedekar and to so many others who shaped the way I think about building and leading. I’m leaving Meta with a much better understanding of both — and the confidence that, for everything I don’t know yet, I can figure it out. Thank you, Meta. 💙
-
Nico Lopez reacted on thisProud doesn't quite cover it. This one started as an idea a few months ago — and today it's real and in people's hands. Getting something from an early concept to launch takes conviction, timing, and above all a team willing to build it with you. I'm grateful for the people who believed in it and brought it to life. Can't wait for you to try it. 🚀 #SpatialComputing #AgenticAI #XR #MetaQuest #OpenXR #AIAgents #InnovationNico Lopez reacted on thisThe agentic loop for XR development is now closed. Today we're shipping Meta XR Operator — a new capability in the Meta XR Core SDK (v205) that gives MCP-compatible AI agents full spatial context and control over running VR and MR applications. Until now, AI agents could reason about code and assets, but couldn't perceive or act inside a live XR session. Meta XR Operator changes that. Any MCP-compatible agent can now: • Query XR session state and per-frame data • Read and set head/controller poses and input • Capture composited physical + virtual views • Traverse the scene graph — rooms, walls, UI canvases, and interactables • Invoke custom C# or OpenXR tools you define It integrates with any OpenXR-based Quest project (e.g., Unity OpenXR) with zero code changes. For the fastest iteration, pair it with Meta XR Simulator — no hardware in the loop. It also runs on Meta Quest devices for on-device validation. This unlocks AI-driven development workflows: automated testing, validation, and iteration — all without a human in the headset. Get started: 🔗 Unity: https://bit.ly/3Tpofw0 🔗 General OpenXR: https://bit.ly/4yShSSf #XR #VR #MR #AI #OpenXR #SpatialComputing #DevTools #MetaQuest #Unity
-
Nico Lopez reacted on thisI've taken a little time to sit with the news before posting, but I'm ready to share and start the next chapter! After 6+ years at Meta, my role was impacted alongside many incredibly talented colleagues. I’m deeply grateful for everything I had the opportunity to build and experience during my time there. From joining at the start of 2020 (right at the beginning of Covid!) to growing alongside such a thoughtful and talented community - it’s been one of the most meaningful chapters of my career. I’ve had the privilege of designing work I truly cared about and build genuine friendships with the people I worked with. I feel very lucky to have crossed paths with so many amazing folks. For what's next, I’ll start actively exploring for new opportunities. If you know of any open roles or would just like to connect, I’d love to chat. And a sincere thank you to this community for any support. It truly means a lot! 🙏
-
Nico Lopez liked thisNico Lopez liked this🤯 It’s been 10 days since I posted my progress on Arcade Hoops 🏀 Since then, I’ve added 50+ commits and a variety of new features: - New onboarding flow with tabletop and floating placement - Mini tutorial to teach the core mechanic - Updated basketball with shader-generated lines - Court and basketball impact shader - Sound effects for scoring, bouncing, and achievements, also 3 new music tracks - Full PSDK integration with 1 leaderboard and 10+ achievements - Achievement toast popup when players unlock one - Persistent Spatial Anchors, so court placement restores after closing the app - Debug area to restart onboarding, toggle aiming lines, and view in-game logs - Improved basketball bounce physics, it feels much more realistic now - Lastly, A LOT of additional fixes & improvements! 💡What’s next? - Creating store metadata, screenshots, trailer, and store details - Then submitting the game 🚀 most likely next week! #meta #xr #ai
-
Nico Lopez liked thisNico Lopez liked thisOk, this GPT Image 2.0 model is seriously impressive. The amount of detail it added when I gave it this prompt: “𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗲 𝗮 𝗨𝗻𝗶𝘁𝘆 𝗶𝗻𝘁𝗲𝗿𝗳𝗮𝗰𝗲, 𝗽𝘂𝘁 𝗺𝗲 𝗶𝗻𝘀𝗶𝗱𝗲 𝘁𝗵𝗲 𝗨𝗻𝗶𝘁𝘆 𝘀𝗰𝗲𝗻𝗲, 𝗮𝗻𝗱 𝗵𝗮𝘃𝗲 𝗺𝗲 𝗿𝗲𝗮𝗰𝗵 𝘁𝗼 𝗼𝗻𝗲 𝗼𝗳 𝘁𝗵𝗲 𝗴𝗮𝗺𝗲 𝗼𝗯𝗷𝗲𝗰𝘁𝘀.” ℹ️ My only note here is that the GPT Image 2.0 model goes rogue sometimes. I kept giving it reference images of how I look, but it sometimes worked and sometimes didn’t. But regardless of its impressive level of detail, especially with user interfaces! #ai #gptimage2
-
Nico Lopez liked thisNico Lopez liked thisAbout 4 days into my upcoming VR/MR YouTube video, and honestly extremely excited to show you how to build a full game from the ground up by using Agentic Tools such as Unity AI, MCPs, Meta MCP extensions, hzdb, and more! 💡If everything goes as planned, this should be out tomorrow or the next day!
Experience & Education
Publications
-
Quasi-fortunate Numbers
Online Encyclopedia of Integer Sequences
See publicationThe Quasi-fortunate numbers are the Least prime m such that either prime(n)# - m is prime or prime(n)# + m is prime, where p# denotes the product of all primes <= p. This set of numbers is useful to find large prime numbers near primorials, and can be used to create algorithms for public-key cryptography like the ones used in the RSA.
Courses
-
Algorithm Design and Analysis
354
-
Computer Networking
330
-
Discrete Math (Python)
136
-
Intro to Math Research (Algorithmic Number Theory)
370
-
Machine Learning
-
-
Modern Physics
304
-
Object Oriented Programming
128
-
Sets, Logic, and Recursion
135
Projects
-
Caption_it, mobile app
- Present
See projectCaoption_it is the first mobile meme making game where you supply the punchlines for pictures and videos that your friends upload. It was programmed in Swift and the Android SDK, and will soon be available in the AppStore.
-
A Primorial Approach for Generating Primes
-
See projectGenerating primes quickly of a given bit size has important applications in multiple fields. For example, RSA, the first public-key cryptosystem, generates 4000-bit primes to secure your data. There are multiple algorithms that rely on generating a random prime within a given bit length and then testing for primality with a strong pseudo-primality test. Currently, this algorithm is characterized by being inconsistent in the time it takes to compute. My goal was to create a computationally…
Generating primes quickly of a given bit size has important applications in multiple fields. For example, RSA, the first public-key cryptosystem, generates 4000-bit primes to secure your data. There are multiple algorithms that rely on generating a random prime within a given bit length and then testing for primality with a strong pseudo-primality test. Currently, this algorithm is characterized by being inconsistent in the time it takes to compute. My goal was to create a computationally consistent algorithm that could be used to find large prime numbers. The algorithm was found to be 91% faster than the standard method when generating primes up to 730 bits with my Primorial Approach.
Honors & Awards
-
Best Hardware Hack at HackHarvard
Harvard
Built a dog training robot, called train-o, that identifies your pet’s pose using OpenCV and Reinforces behavior by dispensing treats at Harvard's Major League Hacking (MLH) event. This project was also featured as part of DevPosts top 5 worldwide weekly hacks. Read more: https://devpost.com/software/train-o-9n85ez
-
McLean County Outstanding STEM Student
McLean County Chamber of Commerce
First-year computer science and physics double major David “Nico” Lopez '21 brings the fun – and the funds – with a $1,000 “Outstanding College Student” scholarship awarded at the McLean County STEM Gala on Wednesday, April 4.
News Article: https://bit.ly/2qIaiJa
-
Illinois Wesleyan Entrepreneurial Fellowship Grant
-
A five-thousand-dollar grant to develop a Caption_it, the first online competitive meme-making game available on Android and IOS.
Languages
-
Spanish
Native or bilingual proficiency
-
English
Native or bilingual proficiency
Other similar profiles
Explore more posts
-
Jonathan Fulton
AI moved fast again last week — from OpenAI reorganizing its ChatGPT team and hiring the makers of Xcode’s coding assistant, to DeepMind’s breakthrough in protein interactions, Amazon’s new AI chip, and the EU’s push for transparency rules. I pulled together the 30 most important stories into one place. https://lnkd.in/dBK5JaHF
4
-
Satishkumar Dhule
Last Tuesday, Deliveroo's rider-switch Rails endpoint spiked to 4s latency and intermittent 503s. The flame graphs told the real story and became our speed playbook. Technical context: Flame graphs visualize time across call stacks using sampling profilers. In Rails, latency often stems from DB calls, serialization, or middleware. This makes root causes visible to newcomers. Recent tech context includes AI agents and RAG patterns (LangGraph, Claude 3.5, GPT-4o, Gemini 2.0) guiding performance analyses. KEY INSIGHTS: 🔍 Flame graphs reveal hot stacks and where wall time sits. ⚡ Latency on the hot path fell from 4s to ~800ms after fixes. ───────────────────────── 🔗 Read the full article: https://lnkd.in/ghcREzpZ 🎯 Practice interview questions: https://lnkd.in/gmy5drNw #performancetesting #cpuprofiling #memoryprofiling #flamegraphs #latency
1
1 Comment -
Hiro Yamada
I’ve been asking engineer friends in SF: “How much more productive are you with AI tools like Cursor and Claude Code?” The answers are all over the place: 👉 Some say ~10% gains. 👉 Others claim 10x. And then there’s the METR study suggesting senior engineers were actually **19% slower** on open-source repos with AI. So where’s the truth? At First Mate Technologies, we’ve been using Cursor and Codex internally. Anecdotally, our team is seeing ~20–30% productivity gains. 💡 Is that in line with what you’re seeing? 💡 Or are we missing a bigger lever? I hear companies are standing up dedicated AI Developer Experience teams to maximize these gains. We don’t have that (yet!)—but I’m wondering if that could be the path to getting from 20–30% to something much larger. Curious to hear: how are you setting up your eng teams and processes to take advantage of AI?
24
1 Comment
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content