NimbleEdge

NimbleEdge + Microsoft: Foundry Local is now on Android. With DeliteAI & Foundry Local, real on-device AI is available — reasoning, context, tool-calling, all running locally. Learn what this unlocks for on-device AI: nimbleedge.com/blog/foundry-l… #OnDeviceAI #FoundryLocal #DeliteAI #NimbleEdge

We are at NeurIPS in San Diego! Come and learn about "Reasoning at the Edge" in the session by our CEO @AKAvkkhare organized by the @PyTorch foundation in their amazing workshop on "Edge AI". Read more: neurips.cc/virtual/2025/l…

Looking forward to exciting sessions on Agentic and Edge AI from the @PyTorch contributors at NeurIPS. Catch our CEO @AKAvkkhare for his talk on “Reasoning at the Edge” with focus on Contextual Sparsity and Streaming First Infrastructure for the Edge.

We’re thrilled to share that NimbleEdge will be at NeurIPS 2025 (San Diego, Dec 2-7). Our co-founder & CEO @AKAvkkhare will be speaking on Dec 3 about “Realtime reasoning at the edge” — and how on-device AI can unlock privacy, speed, and scale. We’re excited to connect with researchers, practitioners, and fellow AI-builders working on edge AI, compiler-toolchain optimizations (quantization, pruning, sparsity), real-time stream based agentic infrastructure and distributed on-device AI. Whether you’re curious about on-device models or want to discuss building AI-native apps for billions of devices — let’s meet. Reach out if you’ll be at NeurIPS. Learn more & check out the full session details: neurips.cc/virtual/2025/l… #NeurIPS2025 #EdgeAI #OnDeviceAI #AIAtTheEdge #NimbleEdge

It's Saturday but Indian tech ain't slowing down. Here's the last 24 hours at a glance: 1. @mayaresearch_ai's Maya1 is on Fal 2. @games_heads showcased Raji: Kaliyuga 3. @PrimeToolings is building MPADS boosters 4. @NimbleEdgeInc partnered with Microsoft India, let's gooo!

PhonePe one of the largest digital payments platform in the world with over 600M (yes Million) is the first enterprise to build with the Foundry Local stack. Glad to be part of this journey to make this possible with Microsoft!

Foundry local for Android announced at #Ignite by Raji Rajagopalan. It’s been amazing to be a key partner with Microsoft in bringing on-device AI to millions of devices on Android!

It's amazing to be at Ignite and meeting the Foundry Local team! Even more excited about the announcements later this week. Stay tuned!! Ignite conference is huge and just walking/going through the lines is half the time spent...and of course you run into some protesters along the way....:P

Our CTO @nrjpoddar will be at Ignite! Come and meet us to talk about on-device AI and how Foundry Local & DeliteAI are helping shape the landscape of on-device intelligence.

We're thrilled to recap our experience at @PyTorch Conference 2025 in San Francisco! As part of the NimbleEdge team, we were honored to be selected for the Startup Showcase, standing alongside incredible AI-first companies. A major highlight was presenting our "Speedy Neurons With Dynamic Sparsity for Fast, Lean LLMs" poster series , showcasing how on-device inference can be both high-performant and resource-efficient. A huge thank you to the PyTorch team, and a special shout-out to industry leaders like Joseph Spisak, Matt White, and the ExecuTorch community for their incredible work. We were particularly proud to be highlighted by the ExecuTorch team during their 1.0 launch as one of their success stories! It's been an amazing journey collaborating with this community, and we're excited for what's next. We gained invaluable insights into inference runtimes, edge deployment, and sparse model optimization - all directly aligned with our mission to scale AI for billions of users. Keep an eye on our space – we're accelerating on-device conversational AI and look forward to sharing our next updates soon! Read full recap here: nimbleedge.com/blog/pytorch-2… #AI #MachineLearning #OnDeviceAI #EdgeAI #PyTorchCon #StartupShowcase

We had an amazing time talking about "Speedy Neurons With Dynamic Sparsity for Fast, Lean LLMs" in PyTorch Poster Session - exciting end to the first day of the conference.

Congratulations to the ExecuTorch team for the v1.0 release. We are excited to see what we can build on-device with ExecuTorch and grateful to be part of this journey from early days. Read more about ExecuTorch and DeliteAI on the ExecuTorch success stories blog here: docs.pytorch.org/executorch/mai…

In 2025, fast + personal + private is the baseline for great product experiences. Our new blog breaks down how privacy-first, low-latency models are making that possible and how NimbleEdge helps deploy them at scale. ➡️Real-time on-device understanding ➡️Sub-100ms inference ➡️Data stays on the device Read here → nimbleedge.com/blog/personal-… #OnDeviceAI #EdgeAI #PrivacyFirst #AIProductDesign

We’re pleased to share that our team at NimbleEdge was part of the “Startup Showcase” at PyTorch Conference 2025—one of only 11 early-stage AI startups selected to pitch live to a panel of industry leaders. Our CEO Varun Khare, presented our vision of making "AI for UI" while making edge-AI accessible and scalable, and we had rich conversations with peers, investors and developers about how NimbleEdge’s platform addresses real-world deployment challenges. We’re grateful for the opportunity to engage in this vibrant community and excited for the main event to kickoff where we will be presenting a poster session on "Speedy Neurons for Faster LLMs"

Excited to share that NimbleEdge AI is joining the "Edge AI Hackathon" hosted by @Meta, @Qualcomm & @github on Oct 20 in San Francisco! Participants will use ExecuTorch, Snapdragon hardware & GitHub Copilot, with mentorship from industry experts, workshops, and hands-on sessions. At NimbleEdge, we’re excited to showcase privacy-first, mobile-first AI solutions and collaborate with the brightest minds in edge and on-device AI. Register here - cerebralvalley.ai/e/executorch-h…

We’re thrilled to see our CEO - @AKAvkkhare, representing NimbleEdge at @GITEX_GLOBAL 2025 in Dubai which is the world’s most established and iconic tech exhibition. GITEX Global continues to be an incredible platform for discovering cutting-edge technologies, connecting with global innovators and exploring how the next wave of AI is being built. Exciting conversations, valuable insights and plenty of inspiration to take forward as we continue building the future of privacy-first on-device AI. #GITEXGlobal #NimbleEdgeAI #OnDeviceAI #AIInnovation

We’re excited to share that NimbleEdge AI will be presenting at the Poster Session during the @PyTorch Conference 2025, taking place on October 22–23 in San Francisco as part of Open Source AI Week. The PyTorch Conference brings together developers, researchers, and innovators from around the world to explore the latest breakthroughs shaping the future of open-source AI. Our poster, “Speedy Neurons With Dynamic Sparsity for Fast, Lean LLMs,” will showcase how we’re rethinking model efficiency, enabling large language models to run faster and lighter through adaptive neuron activation. If you’re attending, drop by our session to discuss the next frontier of efficient, on-device AI and how we can make powerful models truly personal. #PyTorch #Startup #NimbleEdge #AI #PrivateAI