About
Amar Krishna is a visionary entrepreneur and innovative leader dedicated to developing AI…
Articles by Amar
Activity
16K followers
-
Amar Krishna shared this𝗜'𝗺 𝗶𝗻𝗰𝗿𝗲𝗱𝗶𝗯𝗹𝘆 𝗲𝘅𝗰𝗶𝘁𝗲𝗱 𝘁𝗼 𝘀𝗵𝗮𝗿𝗲 𝘁𝗵𝗮𝘁 Giovanni Coglitore 𝘄𝗶𝗹𝗹 𝗯𝗲 𝗷𝗼𝗶𝗻𝗶𝗻𝗴 𝘁𝗵𝗲 𝗕𝗼𝗮𝗿𝗱 𝗼𝗳 𝗗𝗶𝗿𝗲𝗰𝘁𝗼𝗿𝘀 𝗼𝗳 Matrice AI. Back in 2018, when I moved to Silicon Valley from Northwestern, I was fascinated by the people who built the infrastructure behind the internet. They weren't household names, but they built the systems that powered Google, Facebook, and the cloud era we all take for granted today. Gio co-founded Rackable Systems, one of the foundational infrastructure companies and took it public (later acquiring SGI and taking latter's name), helped build Google's early production infrastructure, and later led hardware engineering at Facebook. His work helped shape the way modern data centers are built. As we build what we believe is the next infrastructure layer—sovereign vision intelligence—having someone with Gio's experience, perspective, and belief in our mission join our Board is both humbling and exciting. Welcome aboard, Gio. I'm looking forward to building together and learning from you. Matrice AI Dipendra Jha, Ph.D. Reda Al-Bahrani
-
Amar Krishna shared thisOh loving the new look Matrice AI. Sovereign Vision Intelligence Infrastructure is super exciting 👏
-
Amar Krishna shared thisHappy 250th America 🇺🇸. A country built on an idea that still asks every generation to do the work: build better risk more protect freedom expand opportunity keep moving forward Grateful to build here. Happy 4th of July 🇺🇸 Matrice AI
-
Amar Krishna posted this𝗢𝘄𝗻𝗶𝗻𝗴 𝗰𝗮𝗺𝗲𝗿𝗮𝘀 𝗶𝘀 𝗻𝗼𝘁 𝘁𝗵𝗲 𝘀𝗮𝗺𝗲 𝗮𝘀 𝗼𝘄𝗻𝗶𝗻𝗴 𝗶𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲. That is the shift happening now. Airports, factories, hospitals, malls, energy sites, and governments already have camera networks. The next question is: Can they run AI on that visual data securely, locally, and under their own control? 𝗧𝗵𝗮𝘁 𝗶𝘀 𝘄𝗵𝗮𝘁 𝘀𝗼𝘃𝗲𝗿𝗲𝗶𝗴𝗻 𝘃𝗶𝘀𝗶𝗼𝗻 𝗶𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲 𝗺𝗲𝗮𝗻𝘀. Not more cameras. A smarter, safer, customer-controlled intelligence layer on top of the infrastructure that already exists.
-
Amar Krishna posted thisIf Matrice AI works, the world doesn’t just have more cameras. It has smarter physical infrastructure. The cameras already installed across hospitals, airports, malls, factories, schools, data centers, transit systems, and energy sites stop being passive recorders. They become part of a real-time intelligence layer. Security teams don’t have to stare at hundreds of screens hoping to catch something. Operators don’t have to wait until after an incident to understand what happened. And governments and enterprises don’t have to send sensitive video to someone else’s cloud just to use AI. They can run vision intelligence inside environments they control fully secure— across existing cameras, existing VMS, and existing infrastructure. That is the world we are building toward at Matrice AI.
-
Amar Krishna shared thisPreach :)Amar Krishna shared thisFocus on what you can control. Build something. Anything. A product. Yourself. A family. Your community. A team. Relationships. Or help someone else build theirs. Don’t complain. Don’t play the victim. Ever. What are you building? #PlayNiceButWin
-
Amar Krishna shared thisYou gotta keep at it, okay to be delusional. Congrats SpaceX
-
Amar Krishna posted this𝗘𝘃𝗲𝗿𝘆 𝘁𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝘆 𝗵𝗲𝗮𝗱 𝗜 𝘁𝗮𝗹𝗸 𝘁𝗼 𝗰𝗼𝗺𝗲𝘀 𝗶𝗻 𝘁𝗵𝗶𝗻𝗸𝗶𝗻𝗴 𝘁𝗵𝗶𝘀 𝗶𝘀 𝗮 𝗺𝗼𝗱𝗲𝗹 𝗽𝗿𝗼𝗯𝗹𝗲𝗺. 𝗜 𝗮𝗰𝘂𝘁𝗮𝗹𝗹𝘆 𝗵𝗮𝗱 𝗮 𝗩𝗖 𝘁𝗲𝗹𝗹 𝗺𝗲 "𝗖𝗩 𝗶𝘀 𝘀𝗼𝗹𝘃𝗲𝗱" :) Better detection. Faster inference. More accurate classification. Find the right model and you're done. It's not a model problem. Models are a commodity now. You can get a decent detection model for almost anything in a weekend. 𝗧𝗵𝗲 𝗿𝗲𝗮𝗹 𝗽𝗿𝗼𝗯𝗹𝗲𝗺 𝗶𝘀 𝗶𝗻𝗳𝗿𝗮𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲. 𝗛𝗼𝘄 𝗱𝗼 𝘆𝗼𝘂 𝗿𝘂𝗻 𝟭𝟬 𝗺𝗼𝗱𝗲𝗹𝘀 𝗮𝗰𝗿𝗼𝘀𝘀 𝟭𝟬,𝟬𝟬𝟬 𝗰𝗮𝗺𝗲𝗿𝗮𝘀 𝘀𝗶𝗺𝘂𝗹𝘁𝗮𝗻𝗲𝗼𝘂𝘀𝗹𝘆? 𝗛𝗼𝘄 𝗱𝗼 𝘆𝗼𝘂 𝗼𝗿𝗰𝗵𝗲𝘀𝘁𝗿𝗮𝘁𝗲 𝗰𝗼𝗺𝗽𝘂𝘁𝗲 𝘀𝗼 𝗰𝗿𝗶𝘁𝗶𝗰𝗮𝗹 𝘀𝗮𝗳𝗲𝘁𝘆 𝗮𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻𝘀 𝗻𝗲𝘃𝗲𝗿 𝗹𝗼𝘀𝗲 𝗽𝗿𝗶𝗼𝗿𝗶𝘁𝘆? 𝗛𝗼𝘄 𝗱𝗼 𝘆𝗼𝘂 𝘂𝗽𝗱𝗮𝘁𝗲 𝗺𝗼𝗱𝗲𝗹𝘀 𝗮𝗰𝗿𝗼𝘀𝘀 𝗵𝘂𝗻𝗱𝗿𝗲𝗱𝘀 𝗼𝗳 𝘀𝗶𝘁𝗲𝘀 𝘄𝗶𝘁𝗵𝗼𝘂𝘁 𝗮 𝘀𝗶𝗻𝗴𝗹𝗲 𝗲𝗻𝗴𝗶𝗻𝗲𝗲𝗿 𝗼𝗻 𝘁𝗵𝗲 𝗴𝗿𝗼𝘂𝗻𝗱? 𝗛𝗼𝘄 𝗱𝗼 𝘆𝗼𝘂 𝗽𝗿𝗼𝘃𝗲 — 𝗰𝗿𝘆𝗽𝘁𝗼𝗴𝗿𝗮𝗽𝗵𝗶𝗰𝗮𝗹𝗹𝘆 — 𝘁𝗵𝗮𝘁 𝘁𝗵𝗲 𝗶𝗻𝗳𝗲𝗿𝗲𝗻𝗰𝗲 𝗿𝗮𝗻 𝗰𝗼𝗿𝗿𝗲𝗰𝘁𝗹𝘆 𝗮𝗻𝗱 𝗻𝗼𝗯𝗼𝗱𝘆 𝘁𝗮𝗺𝗽𝗲𝗿𝗲𝗱 𝘄𝗶𝘁𝗵 𝗶𝘁? That's what breaks every existing stack. Not the model. The infrastructure underneath it. Verkada solved the camera. Genetec solved the VMS. Nobody solved the compute and intelligence layer that sits between the camera and the decision. That's what we built at Matrice AI. GPU-first edge inference. Multi-model orchestration. Confidential compute with cryptographic attestation. Continuous learning without cloud dependency. One stack, ground up, built for the scale and trust requirements of critical infrastructure. QLAD Ax3.ai Inc
-
Amar Krishna posted thisI told myself I wouldn't post about this until it felt real. It feels real. 𝗪𝗲 𝗵𝗮𝘃𝗲𝗻'𝘁 𝗺𝗮𝗱𝗲 𝗮 𝘀𝗶𝗻𝗴𝗹𝗲 𝗰𝗼𝗹𝗱 𝗰𝗮𝗹𝗹. 𝗘𝘃𝗲𝗿. But somehow we're deep in conversations with a major energy company in the Middle East, one of the largest mall operators in Southeast Asia, a police force, hospital networks, transit agencies, government bodies, and just got off a call with our partners whose customers — big telcos, infrastructure operators — have been asking for exactly what we built. Qatar. Indonesia. Philippines. UK. USA. All inbound. All through partners. Nobody told me this would happen. Honestly I didn't expect it this fast. Building a startup is mostly hard and lonely and uncertain. But every once in a while something happens that makes you think — okay, we're onto something real here. This is one of those moments. #SovereignAI #VisionAI #StartupLife #Founder #Matrice
-
Amar Krishna liked thisAmar Krishna liked thisThis one page changed my life. 42 years ago today, it convinced my parents I shouldn't go back to college. I started by upgrading PCs from a dorm room. Today Dell Technologies is helping build the infrastructure that powers AI, from the edge to some of the world's largest AI factories. You never know which small decision will change everything. Keep building 🚀🙏 #PlayNiceButWin
-
Amar Krishna liked thisAmar Krishna liked thisYou know you've made it in life when folks are faking you. Don't fall for it. This ain't me. Stay alert WhatsApp
-
Amar Krishna liked thisAmar Krishna liked thisRunpod is about to look more like a growth at all costs business. But it's a whole lot more nuanced than that. In the next few quarters, we're going to be: - Hiring faster - Investing more in product - Investing more in capacity etc. Here's what the outside can't see: there's a version of growth at all costs that is a deliberate decision, and a version that is a reflex. They both look similar from the outside, but very different internally. The reflex version is... rash: burn money, grow at any price, figure out the economics later. That's a recipe for disaster, and I've watched it take down companies that had every advantage. The deliberate version starts from the outcome. We set out by saying "here's the company we're convinced Runpod needs to become in three, five, ten years." We have set goals, and it becomes a game of getting there on target. Getting there means spending ahead of revenue and, for a stretch, looking like we've stopped caring about efficiency. (The spending might be identical line for line to the reflex version!) But the mode of operation is completely different, because one of them knows exactly what it's driving toward and one of them is hoping. The reason I'm comfortable turning the dial now is, frankly, we've spent years earning the right to. We proved, to ourselves and everyone watching, that we can run a disciplined company. Now it's time to push the throttle and move Runpod into it's next stage.
-
Amar Krishna liked thisAmar Krishna liked thisI was honored to meet HH Sheikh Mohammed bin Rashid Al Maktoum, Vice President and Prime Minister of the UAE and Ruler of Dubai, in the city of London. It is always inspiring to see His Highness among the people, as his presence reflects the spirit of a true leader who instills pride and inspires everyone through his values and actions.
-
Amar Krishna liked thisAmar Krishna liked thisWTF in bold. a $32 billion startup. nobody know what it does, but not to worry. we know its AI is “safe” and the founder is a “genius”. nvidia gets to capitalize its investment while booking revenue. how long can this madness last???Nvidia bets $5bn on Ilya Sutskever’s AI breakthroughNvidia bets $5bn on Ilya Sutskever’s AI breakthrough
-
Amar Krishna liked thisAmar Krishna liked thisThe PCAC hearing on peptides wasn’t really about if patients will pursue these options. They will. The real question: does the FDA push patients toward a regulated healthcare system, or an unregulated marketplace with zero guardrails? As American Academy of Peptide Medicine put it: this was never peptide care or no peptide care. It’s regulated or unregulated. Regulated means: • Individual evaluation by a licensed provider • Compounding by a licensed pharmacist • Verified API identity and purity (COAs) • Controlled, tailored dosing Prohibition doesn’t kill demand. It kills oversight. If patient well-being is the real priority, keep this inside a regulated, accountable system, not push it out where nobody’s watching. Compounding is not a crime. It’s personalized medicine, done right. Big win today! Peptides are coming back!
-
Amar Krishna liked thisThis is an extremely exciting milestone for the Hailo team and for me personally. After more than nine years as an independent company, we have signed a definitive agreement to join Microchip Technology Inc. and advance our vision of AI at the edge—a vision that is more relevant today than ever. The journey was never easy, but it was worth every moment. It was frustrating, satisfying, funny, nerve-racking, and illuminating—often all at once. I am very proud of everything we have achieved over the years, and deeply grateful to everyone who helped us build this vision and make it a reality: my co-founders and colleagues at Hailo, our investors, customers, suppliers, and partners—and, most of all, my family for their unwavering support throughout the journey. Many good things lie ahead in this next chapter of our story, and I look forward to continuing to work with this amazing group of talented people!Amar Krishna liked thisToday, we are pleased to share that Microchip Technology Inc. has signed a definitive agreement to acquire Hailo. Hailo’s mission has always been to deliver industry-leading Edge AI processors that enable our customers to bring innovative AI solutions to market. As we look toward our next chapter, we are excited about the opportunities this combination will create for our customers, partners, and employees. Microchip shares our commitment to innovation, technical excellence, and customer success. We have been particularly impressed by its global reach, long-term customer relationships, manufacturing capabilities, quality standards, and extensive technical support organization. We remain fully committed to supporting the incredible community of builders using our products to make Edge AI a reality. We look forward to continuing this journey together. Stay tuned—we are just getting started! To read more: https://lnkd.in/danJZwkVMicrochip Technology Signs Definitive Agreement to Acquire HailoMicrochip Technology Signs Definitive Agreement to Acquire Hailo
-
Amar Krishna liked thisAmar Krishna liked thisExcited to share that I'm currently consulting with Everest Labs, a physical AI company building intelligent systems for material separation - with recycling as our first application. Great to be contributing to a mission-driven team at the intersection of AI and sustainability, applying cutting-edge technology to real-world environmental challenges. More to come!
-
Amar Krishna liked thisIt is a great pleasure to join the Reticore Labs team as an advisor. Mustafa Yazgan, Altan Odabasi and Tinu Thomas are an amazing team with great talent, experience and insight to take AI-driven EDA to the next level.Amar Krishna liked thisThrilled to welcome Aykut Dengi to the Reticore Labs advisory team! Aykut's career spans the full arc of silicon innovation. He began in the semiconductor industry at Motorola/NXP, then helped pioneer AI for performance-driven interactive layout at Neolinear, continuing through its acquisition by Cadence. At Arizona State University, he researched digital design for high-performance, energy-efficient, and distributed inference. Most recently, as CEO and CSO of RadiusAI, he brought vision AI to retail at scale, and he advises CEMWorks on large-capacity electromagnetic simulation for high-performance digital designs and multi-chip solutions. An innovator with over 40 patents, Aykut bridges the gaps between silicon, design automation, and AI. At Reticore, we're building AI for the design automation of silicon AI systems - and few people have lived at that intersection longer than Aykut. His deep domain expertise and strategic guidance will be invaluable as we accelerate our roadmap and scale our technology. Please join us in giving Aykut Dengi a warm welcome to the Reticore Labs advisory team! #EDA #Semiconductors #AI #ChipDesign #DesignAutomation #HardwareAcceleration #StartupUpdate #ReticoreLabs
Experience & Education
-
Matrice.ai
View Amar’s full experience
See their title, tenure and more.
or
Publications
-
Polarity Trend Analysis of Public Sentiment on YouTube
Amar Krishna, Joseph Zambreno, Sandeep Krishnan
Projects
-
Topic modeling and document classification using LDA, pLDA, Pachinko Allocation Modeling, LDA with gibbs sampling.
-
Investigation of Traffic Volume vs. Route Suggestion in Google Maps
There are various map services available these days. The major information they provide include the road information (construction, blockage, etc.), the live traffic, and route suggestions based on those aforementioned factors. Routes are being suggested given the origin and destination of trip, as well as the trip setoff time. The travel means also matter but in this project we only focus on driving (excluding walking, bicycling, taking a bus, etc.). However what they do not disclose is the…
There are various map services available these days. The major information they provide include the road information (construction, blockage, etc.), the live traffic, and route suggestions based on those aforementioned factors. Routes are being suggested given the origin and destination of trip, as well as the trip setoff time. The travel means also matter but in this project we only focus on driving (excluding walking, bicycling, taking a bus, etc.). However what they do not disclose is the mechanism behind route suggesting. We are interested to see if there exist relationships, and what relationships are there, among different routes suggested from different times, as well as between routes and traffic itself. Furthermore, we are interested in the blackbox algorithm they might be using, and attempt to use a reversed engineering way to figure it out. A good understanding of the mechanism of route suggesting can help us make better decisions in the future.
-
Sentiment Analysis of Movie Reviews
The goals of this project were
• Classify the sentiment of movie reviews from rotten tomatoes and YouTube comments by correlating the movie ratings from both datasets and evaluating the proximity of their values.
• Estimating polarity trends by calculating the polarity of each review using the Naive Bayes classification method. We also used R for our statistical analysis (i.e. the polarity values).
Languages
-
Spanish
Elementary proficiency
-
English
-
-
Hindi
-
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