Co-founder and CTO at Nexa AI, Industrial Veteran from Google & Amazon, and Stanford alumni. Committed to lifelong learning and advancing AI technology.Joined October 2021
Run llama3.2-3B on Snapdragon NPU with 2X performance vs others, in one line of code:
sdk.nexa.ai/model/Llama3.2…
nexa infer NexaAI/Llama3.2-3B-NPU-Turbo
Run llama3.2-3B on Snapdragon NPU with 2X performance vs others, in one line of code:
sdk.nexa.ai/model/Llama3.2…
nexa infer NexaAI/Llama3.2-3B-NPU-Turbo
Most teams trying to get their models onto NPUs today face the same roadblocks:
❌ Conversions have to be done case by case, with tons of manual work
❌ Support is usually limited to only a few model families (Llama, Qwen, etc.)
That’s the pain point we set out to solve.
With…
Most teams trying to get their models onto NPUs today face the same roadblocks:
❌ Conversions have to be done case by case, with tons of manual work
❌ Support is usually limited to only a few model families (Llama, Qwen, etc.)
That’s the pain point we set out to solve.
With…
We have successfully enabled multimodal (image) support for Gemma-3n in Nexa SDK — a highly requested capability that is currently not available in llama.cpp or Ollama so far, where Gemma-3n remains text-only. @GoogleDeepMind @Google#AI#Multimodal#Gemma3n#Innovation
We have successfully enabled multimodal (image) support for Gemma-3n in Nexa SDK — a highly requested capability that is currently not available in llama.cpp or Ollama so far, where Gemma-3n remains text-only. @GoogleDeepMind @Google#AI#Multimodal#Gemma3n#Innovation
🚀 We’ve launched OmniNeural — the world’s first multimodal AI model optimized for NPU. With Nexa SDK, devs can run NPU models in just 1 line of code. 🧵
1️⃣ AI Action Assistant
Voice-driven control that actually does things: call, text, email — instantly, offline, and private.…
Yes, @elon@Qualcomm@Snapdragon I couldn’t agree more — at Nexa AI we’re making this future real. We train and deploy AI models to fit on-device, turning every phone into an intelligent edge node. 🚀
Yes, @elon@Qualcomm@Snapdragon I couldn’t agree more — at Nexa AI we’re making this future real. We train and deploy AI models to fit on-device, turning every phone into an intelligent edge node. 🚀 https://t.co/0mbMku1rOu
I truely believe in this vision. At Nexa AI, we build on-device model and inference framework that runs any model on any device, across any backend, see : sdk.nexa.ai
I truely believe in this vision. At Nexa AI, we build on-device model and inference framework that runs any model on any device, across any backend, see : sdk.nexa.ai
How fast can you get @Alibaba_Qwen's Qwen3-1.7B model running on @Snapdragon NPU with nexa-sdk? My speedrun: < 1 min 🚀
And yes—it also runs OmniNeural-4B, YOLOv12, PaddleOCR, and more.
What’s your fastest run?
🚀 With Nexa ML, running CV & multimodal models on-device takes 1 line of code.
Blazing-fast ⚡, power-efficient 🔋 inference on the Snapdragon X Elite NPU — while your CPU/GPU stay free for work & play (PowerPoint, video, gaming 🎮).
🎬 Demos:
📄 PaddleOCR → instant text…
📱Mobile Demo: On a @SamsungMobile S25 Ultra, the user takes a photo of a prescription and asks:
“What’s listed here?” → “Set a reminder for me.”
OmniNeural-4B sees the image, understands the voice, and sends the right command to the reminder app.
➡️ Running fully on the…
My friend Alex build @nexa_ai has been incredible to witness 👏
Today, they launched OmniNeural - a strong multimodal AI that runs on your phone's NPU. Real-time, private, blazing fast. Huge congrats to them! 🎉
#AI#Innovation#Startup
My friend Alex build @nexa_ai has been incredible to witness 👏
Today, they launched OmniNeural - a strong multimodal AI that runs on your phone's NPU. Real-time, private, blazing fast. Huge congrats to them! 🎉
#AI#Innovation#Startup
Almost every new PC, phone, car, and IoT device ships with an NPU built for AI.
Yet most models remain GPU-first — leaving NPUs idle while on-device AI burns battery, lags, and struggles to multitask. Hardware evolved. Software didn’t.
That’s why we built OmniNeural-4B + nexaML:…
Almost every new PC, phone, car, and IoT device ships with an NPU built for AI.
Yet most models remain GPU-first — leaving NPUs idle while on-device AI burns battery, lags, and struggles to multitask. Hardware evolved. Software didn’t.
That’s why we built OmniNeural-4B + nexaML:…
20 Followers 51 FollowingUnlock the Cosmos! Unique Universe NFTs: Explore exclusive digital art, collect rare NFTs & enter a universe of endless creativity.
3 Followers 137 FollowingWeb 3.0 AI execution protocol
AI Layer for Web3 with a goal of onboarding 100 Million users.
Powered by $GTAI token.
Incubated by @Chain_GPT
22 Followers 51 FollowingDigital Coin Bank mints coin NFT's that are an artistic representation of the digital currency focusing on the numizmatic beauty aspect of the coins
18 Followers 42 Following🔥 Memecoins, gemas, gestão de risco, realização de lucro
📊 Transformando experiência em conteúdo
📥 Canal pago com insides e calls no Telegram .
15 Followers 45 FollowingMulti-Media ᵖᶳᵞᶜᶣᵒ Artist ☮️🍄
"True art whispers eternity, hype shouts for just a moment"
🔺https://t.co/dCEdkWAlj9
💧https://t.co/V8kFPVW8Au
8 Followers 144 FollowingNFFT is the next evolution of NFTs
▪️
9112 collectible and fillable Blocks with proof of ownership
stored on the Ethereum blockchain
3 Followers 306 FollowingThe latest news and more from Rolling Stone magazine and https://t.co/qu7A98O95S. Got a tip? Share it here: https://t.co/Vh2uz40Dv1
451K Followers 77 FollowingTensors and neural networks in Python with strong hardware acceleration. PyTorch is an open source project at the Linux Foundation. #PyTorchFoundation
1.2M Followers 279 FollowingWe’re a team of scientists, engineers, ethicists and more, committed to solving intelligence, to advance science and benefit humanity.
18K Followers 1K FollowingVP of DevRel for @GitHub. Previously Executive Director @dotnetfdn and original creator of the @Microsoft org on @GitHub (he/him)
63K Followers 456 FollowingFreelance marketer building TechnoBizzVault, where I help professionals discover modern tools so they can be productive without burning out.
329K Followers 4K FollowingCo-founder of Tiny w/ @_Sparling_. We own @Dribbble, @Serato, @Letterboxd, @AeroPress, and 35+ other wonderful companies. Author of Never Enough.
780 Followers 207 FollowingRobotics Engineer/Founder. Into Multimodal GenAI. Was previously at @microsoft @paypal @jibo @MistyRobotics. Hit me up for a pickleball or squash game anytime!
13K Followers 2K FollowingBuilding Digital PR backlinks for SaaS and other SEO clients
📰 Featured on Forbes, BusinessInsider, Verge, DigitalTrends...
Need backlinks? 💬 DM me!
.
56K Followers 853 FollowingFiguring out AI @allen_ai, open models, RLHF, fine-tuning, etc
Contact via email.
Writes @interconnectsai
Wrote The RLHF Book
Mountain runner
282K Followers 2K FollowingSenior Editor @verge ║ Sign up to Notepad, my weekly newsletter on Microsoft's big bets at https://t.co/KqkAib2CKP ║ Tips? Msg on Signal app: tomwarren.01
9K Followers 6K FollowingProduct Lead @Firebase (Serverless & AI). Working all day, testing AI tools and models all night.
Prev: Microsoft. Opinions are my own
15K Followers 842 FollowingQualcomm Ventures is the venture investment arm of Qualcomm, Inc. Investing in cutting edge, innovative #startups.
San Diego, CA
51K Followers 914 FollowingPresident and CEO of @Qualcomm. Husband and proud father. I share my views about wireless and related technologies. Opinions are mine.
7K Followers 697 FollowingMarketing leader & @Qualcomm CMO telling stories about the next-gen tech that’s changing the world. Sports, pop culture, and @Snapdragon fan. Opinions are mine.
2K Followers 245 FollowingChief Commercial Officer @Qualcomm. Passionate about how connected tech will transform our lives. Music and music tech aficionado. Tweets are my own.
12K Followers 513 Following⚡ Founder & CEO helping you level up | Follow for insights on self-improvement, leadership and AI productivity | 📢 TEDx Speaker | 🔨 T&O Surgeon
54K Followers 1 FollowingWorkflow automation for technical teams to build AI solutions that integrate with any app or API at no-code speed and code flexibility. Open and self-hostable
221K Followers 2K FollowingWhere “imagine if” gets to work. We've helped 500+ companies (like @NotionHQ, @Roblox, @Uber, @Square) take a straighter path from idea to product-market fit.
No recent Favorites. New Favorites will appear here.