About
I'm a data/AI scientist who loves exploring the world of AI and innovation. I feel…
Services
Articles by Steve
Activity
2M followers
-
Steve Nouri shared thisThis saves you hundreds of hours easily. 🏆 AI agents are great but the people who can make businesses out of it are the smartest. This is one of them. To sum it up, it's just like a dedicated CX manager that updates you through iMessage or Slack on all customer-facing interactions. It's called Chatbase. You set it up once, it runs your customer support in the background and it texts you what actually matters while you go about your day: → Summaries of every customer conversation, no digging through the weeds → Patterns you'd never catch: "14 people asked about the same feature today, might be worth building" → What's working, what's not and what to fix Want to know something about your support? You just text it and ask. No dashboards. The feature that got me is Actions. The agent plugs into Stripe, Shopify and HubSpot and resolves requests end to end, from processing refunds to pulling invoices. And when a case needs a human, it preps the resolution and hands it over. If you're running a business without the time or people for support, this is how you scale without hiring. The full setup is in the video. That's how you get hundreds of hours back.
-
Steve Nouri shared thisMassive Shift! Enterprise AI is entering a new phase. #IBMPartner For the last year, most of the conversations have been about models, speed, and adoption. But for CIOs, CTOs, and technology leaders, the deeper question is becoming: Can we innovate with AI while keeping control over data, governance, compliance, and operations? That is why digital sovereignty matters. The future of enterprise AI will not only be about what AI can generate. It will be about whether organizations can prove how it is governed, where it runs, who controls it, and how risk is managed. IBM Sovereign Core is designed around that challenge: helping organizations operate AI-ready environments with control, governance, and verifiable trust : https://lnkd.in/gdrC3-2K AI adoption is accelerating. The winners will be the organizations that can scale it with provable control. Is digital sovereignty or the speed of AI adoption a bigger priority for you right now? Why?
-
Steve Nouri shared thisI think this is an important reality check. I’m less interested in whether AI can describe an image, and more interested in whether it can interpret what an expert actually sees in their daily work. That 7.9–26.5% accuracy range shows me the gap is still huge. I see benchmarks like VIALS as critical because they test something much closer to real scientific reasoning, not just polished demos.Steve Nouri shared thisAI promises to cure disease, but can frontier models interpret the images biotech/pharma scientists handle every day? To evaluate this, we built the ATLAS Visual Life Sciences (VIALS) benchmark: 161 visual interpretation tasks from professional life sciences workflows including gel blots, plasmid maps, molecular structures, and phylogenies. These are the images/results that experiments actually produce, not polished figures from papers or textbooks. We benchmarked today’s 10 leading multimodal models. Opus 5, GPT 5.6, and Gemini 3.7 all fail to crack 27% accuracy, and the other models like Kimi K3 & Grok 4.6 perform much worse. In contrast, PhD-level biomedical scientists find these visual interpretation tasks straightforward. AI that cant interpret such images will have limited utility in life sciences R&D, where such artifacts are central to how scientists reason, communicate, and make research decisions. We hope VIALS drives progress toward models that can
-
Steve Nouri reposted thisFor years, enterprise software was designed to help people do work. The next generation of software is designed to do the work. That's a much bigger shift than most people realize. When people think about AI agents, they usually focus on the model: Which LLM? Which benchmark? Which framework? But after spending time with enterprises deploying AI at scale, I've become convinced that the bottleneck is moving elsewhere. The winning organizations won't necessarily have the smartest models. They'll have the best context. The clearest processes. The strongest governance. And the ability to safely connect AI to real business decisions. That's why I found Oracle's recent announcement particularly interesting. What stood out wasn't another AI assistant. It was the focus on execution. Agents operating within workflows, approvals, permissions, policies, and enterprise guardrails. That's where the real challenge begins. And that's where most organizations still have work to do. If you're exploring agentic AI, Oracle's new Fusion Agentic Applications are worth looking at, not just for the technology itself, but for what they signal about where enterprise AI is heading next. https://lnkd.in/gsgFGZum The conversation is shifting from: "Can AI help people work?" to "Can AI safely move work forward on its own?" #artificialintelligence #oraclepartner #aiagents
-
Steve Nouri reposted thisShadow AI is becoming the new shadow IT. People across companies are already using AI to build apps, automate workflows, and connect tools together. That sounds exciting. Until those apps touch customer data, finance workflows, internal permissions, audit logs, or production systems. Then the question changes. It is no longer: Can AI help people build apps faster? It becomes: Can the enterprise let people build without losing control? That is why I think secure, governed vibe coding is becoming a real category. Superblocks 3.0 is built around this idea: business teams can build production internal apps from a prompt, while IT keeps governance, security, SSO, permissions, audit trails, and policy controls in place. The important part: apps run inside the company’s own AWS account and VPC, so sensitive data stays under enterprise control. And IT can see all of it. Every app, who built it, and what data it touches, in one place. And with multi-agent orchestration, Superblocks routes simpler tasks to cheaper open-source models and harder tasks to frontier models, which keeps token costs down instead of passing them through. The old choice was: Block AI app building. Or open the floodgates. The new option is: Let people build, but make governance the default. That is how shadow AI becomes governed AI. Check out Superblocks: https://lnkd.in/gxnhJmwE Is anyone at your company tracking shadow AI? #superblockspartner
-
Steve Nouri shared thisIf you use Claude, Codex, or Kimi or Manus, you must check this new MCP. I came across a tool called Keemable. They said they’re “an agent browser that learns” and I wanted to know what it does. I research using AI a lot. I send it sources to read. If you do the same for writing, academic studies or planning a new architecture you need AI to be able to view all links in full. What you might NOT know: Even to this day, AI agents like Claude can’t read all webpages. Our human browsers don’t work for them. This is why AI agent browsers are becoming a thing. Might sound complicated but all you need is to connect it to your agents. You can run a command to MCP it. It provides back and full end access to AI agents so they can read everything available and be more accurate with the outputs. It’s a WebQL product. A query layer that lets AI pull structured, relevant results from the long tail of the web rather than relying on general-purpose search built for humans. Your AI can’t do it as efficiently now. It’s one of those things you don’t realize how bad you need it until you try it. Less tokens and more accurate outputs. If your prompts keep failing you, now you know why. They have a free tier that’s more than enough for most users. → https://lnkd.in/grpZacpH
-
Steve Nouri reposted thisAre you building something exciting with open models? I’ve partnered with NVIDIA to give AI builders a shot at a Golden Ticket to #NVIDIAGTC Berlin. 🎫 If you’re building with open models, show us what you’re working on, you could be joining us in Berlin. Your Golden Ticket includes: → Free conference pass → VIP seating for Jensen Huang’s keynote → Exclusive NVIDIA merch → Access to special events How to enter: Share your project below. 📅 Submissions: Aug 18 – Sep 10, 2026 🏆 Winner announced by September 14 👉 Contest details: https://nvda.ws/4hP8W9S Building something worth seeing? "Comment" your project below.👇 Or tag an AI builder who should be in Berlin.
-
Steve Nouri reposted thisI keep coming back to the same point: AI is changing the skill stack faster than most org charts can react. I don’t think the biggest shift is “everyone needs to become more technical.” It’s almost the opposite. I see more value moving toward people who can reason, solve problems, make judgment calls, and use AI well. Problem-solving is still #1. Creativity is rising. Data analytics + AI stay near the top. I find that interesting because I’m seeing the same thing inside companies: AI can absorb more execution, but it still needs humans who know what matters, what good looks like, and what to do next. I wouldn’t bet only on learning another tool. Become better at: thinking + deciding + creating + using AI as leverage. That skill mix is going to matter a lot more than any single platform.
-
Steve Nouri shared thisOkay, this is quite revolutionary. UI design is basically dead. Or very close to it. Here’s what AI can do today ↓ #ElevenAgents #ElevenLabs #ad #ai #ElevenAgentsPartner For literally decades, every piece of software followed the same rule: people had to learn where to click, which menu to open, how to navigate a screen, etc. As apps became more powerful, they also became more complicated. Even simple tasks often required multiple steps and instructions before anything could get done. Thanks to AI, that's no longer how things work. We're moving toward a world where software doesn't need to be navigated at all. You simply talk and get things done through conversation. That's what makes ElevenAgents from ElevenLabs particularly interesting to me. It allows businesses to build intelligent voice and chat agents in minutes using their own knowledge base, SOPs and policies, so every conversation reflects how the company *already* operates. These agents can speak in 70+ languages, connect with tools like Salesforce, HubSpot, Google Calendar and Zendesk, and communicate with natural speech that keeps conversations flowing smoothly. It's a capability trusted by organizations ranging from solo founders to companies where employees at over 75% of Fortune 500 businesses already use ElevenLabs technology. This genuinely feels like a glimpse of how people will interact with technology in the years ahead. If you want to see what that looks like, check ElevenAgents out here: https://lnkd.in/g2u7sf66 How much time do you think people waste every day figuring out where to click? Does it feel like a problem to you?
Steve Nouri commented on a post
14h
Try it here →
https://utm.genai.works/r/ttq12jo
-
Steve Nouri liked thisThis saves you hundreds of hours easily. 🏆 AI agents are great but the people who can make businesses out of it are the smartest. This is one of them. To sum it up, it's just like a dedicated CX manager that updates you through iMessage or Slack on all customer-facing interactions. It's called Chatbase. You set it up once, it runs your customer support in the background and it texts you what actually matters while you go about your day: → Summaries of every customer conversation, no digging through the weeds → Patterns you'd never catch: "14 people asked about the same feature today, might be worth building" → What's working, what's not and what to fix Want to know something about your support? You just text it and ask. No dashboards. The feature that got me is Actions. The agent plugs into Stripe, Shopify and HubSpot and resolves requests end to end, from processing refunds to pulling invoices. And when a case needs a human, it preps the resolution and hands it over. If you're running a business without the time or people for support, this is how you scale without hiring. The full setup is in the video. That's how you get hundreds of hours back.
-
Steve Nouri liked thisSteve Nouri liked thisIs AI now competing with real musicians? Australia decided that fully AI-generated songs can no longer appear on its official charts or be eligible for its music awards. AI can still be used in production, and humans must write the song and perform the lead vocals and main instruments. The move followed controversy around DJ Josh Fawaz's AI-assisted cover of Madonna's "Like a Prayer," which reached No. 1 on the ARIA dance chart and No. 2 overall. The scale behind the decision is the part worth sitting with. A SubmitHub study found nearly 40% of tracks released globally in July 2026 used AI, with 23.2% fully AI-generated. Spotify is responding too, introducing "AI Persona" labels and removing millions of spam tracks. Australia made its decision. Other countries will have to make theirs. Our round is open from $1,000, with up to 22% bonus shares before August 31: https://lnkd.in/d-TnGkAk
-
Steve Nouri liked thisSteve Nouri liked thisThat's pure gold 😯 The engineer who created Claude Code from scratch just released a 28-minute video, and the target: how to write prompts that actually deliver good results. You've seen $300 courses that don't even come close to what he explains in the first 10 minutes. CLAUDE.md files, memory shortcuts, parallel sessions and prompting patterns almost nobody uses. Doesn't matter if you're a developer, just starting out, or have been using Claude for months: this changes the game for you starting today. All in one single video. Completely free. Our round is open from $1,000 with up to 22% bonus shares before August 31: https://lnkd.in/dVKpWyBK
-
Steve Nouri liked thisSteve Nouri liked thisWHOEVER LEAKED THIS HAS ABSOLUTELY NO FEAR It looked like another AI experiment until the setup showed up: → One bot scans the market → Another hunts for setups → A third bot analyzes news and sentiment → Another tracks whales and on-chain activity → Another manages risk and exposure Then a HEAD GROK TABLE BOSS coordinates everything and sends the owner ONLY the trades that really need approval. No team of analysts. No staring at charts all day long. Just 350 Grok bots passing information between themselves 24 hours a day, 7 days a week. And the craziest part: the owner says he barely touches the system anymore. Systems like this only reach the people who could build them when someone puts them in front of them, which is the business GenAI Works runs. We're raising, and the round ends in 3 days. Invest from $1,000 with up to 22% bonus shares before August 31: https://lnkd.in/d5Xu4Dnw
-
Steve Nouri liked thisSteve Nouri liked thisJuggling ChatGPT, Claude, Grok and Gemini for multiple tasks is a living hell. But leveraging them all is a superpower. a16z's latest CIO survey found that 81% of large companies now run 3 or more AI models side by side. A year ago, that number was just 68% (because no single model wins at everything anymore). So people end up with 5 tabs open, 5 logins, 5 bills and honestly no clue which answer to trust. Thats why nexos.ai exists. One browser tab, every major model chatting side by side, so you can compare answers. You can even build a no-code agent trained on your own docs and brand voice, so it handles the repetitive tasks for you every day. It's EU-hosted, fully certified + it never trains on your data. Try nexos.ai with code genai and get 75% OFF using this link https://lnkd.in/ggAic6JV If you had to retire one AI tool today, which would it be?
-
Steve Nouri liked thisSteve Nouri liked thisYour brain is running on just 12 watts right now while processing this sentence. This is not a typo. An AI system would need 2.7 billion watts to do the same thing. The difference comes down to design. In your head, memory and processing happen in the same place. Nothing has to travel back and forth. In a computer, moving data between memory and processor is where most of the energy disappears. Your brain is also lazy on purpose. Most neurons sit quiet most of the time and fire only when something happens. A GPU powers its entire array whether the math needs it or not. That's the gap neuromorphic chips like Intel's Loihi and IBM's NorthPole are trying to close, by copying those two tricks. Evolution had no choice. A brain that burned more energy than the body could find would kill its owner. We built for capability first. Efficiency is the part we're working on now. The companies solving that problem still have to reach a market to matter, and that's the business GenAI Works runs: go-to-market for NVIDIA, Oracle, Google and 300+ AI brands through a 14M+ community, revenue up 2.5x last year. Our round is open. Invest from $1,000 with up to 22% bonus shares before August 31: https://lnkd.in/dq4iKfwr
-
Steve Nouri liked thisSteve Nouri liked thisJapan's new quantum computer uses an American chip Japan switched on Shunkai on August 24, its first full-stack neutral-atom quantum computer. It starts with around 50 qubits and is expected to grow to about 500. The system was built by three partners: - Japan's IMS led the project. - Hitachi built the software. - U.S.-based Infleqtion supplied the quantum processor. Your takeaway: Japan is building national quantum infrastructure and chose the best available technology over an all-domestic build. It also shows how quickly neutral-atom quantum computing is moving from research into real-world systems. Reaching the market early is what decides which of these technologies people actually use, and that's the business GenAI Works runs: connecting NVIDIA, Oracle, Google and 300+ brands with a community of more than 14M people. Now our community can take part in that growth too. Our investment round is open. Invest from $1,000 and receive up to 22% bonus shares before August 31: https://lnkd.in/dRxH36_z
Experience & Education
Volunteer Experience
-
AI Committee Member
ISO - International Organization for Standardization
- 2 years
IT-043 AI committee member of standards Australia contributing to international standards
Honors & Awards
-
Winner of The ICT Professional of the year, Gold Award
ACS - The Professional Association for Australia's ICT sector
"The ICT Professional of the Year is a person who demonstrates excellence in their field of ICT. They are conscious and active in their continuing professional development and display the highest standards of Professional Ethics. They use their technical knowledge to positively to be a role model and influence the workplace professionally, ethically and with the highest standards."
-
Winner of the NASSCOM Innovation Awards 2018
Nasscom
The NASSCOM Student Innovation Awards promote innovation in the field of Information Technology by providing students a platform to display their talent and receive unparalleled exposure to the industry’s experts.
Criteria :
Innovation, Performance, and Potential. Nominees have to demonstrate that their innovation is unique, provides real user benefits and has great worldwide potential. The judges may confer as many or few awards as they agree meets these key criteria. -
1st place at UTS Kaggle competition
-
-
Academic Excellence Award
University of Technology Sydney
Awards are granted on the basis of academic merit
-
Annual Dean's List Award
University of Technology Sydney
The Dean’s List recognizes outstanding academic achievement in postgraduate courses in Engineering and Information Technology.
To be eligible for inclusion on the Dean's List, a student must have:
achieved a minimum weighted average mark (WAM) of 85 for subjects studied in their FEIT course for the previous academic year; and
gained a minimum of 24 credit points of graded subjects in their FEIT course for the previous academic year. -
Winner of the ICT Student of the year
Australian Computer Society
It is a national award, "Awarded to students who have demonstrated academic excellence during their studies, innovation, entrepreneurship and professionalism in undertaking individual and team projects. They have embarked on the journey to become an ICT professional, demonstrating commitment to the profession, extracurricular activities and ethical studies"
Languages
-
English
Native or bilingual proficiency
Other similar profiles
Explore more posts
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore More