CEO at Neurony Solutions | AI Implementation Strategist | Helping Clients Infuse AI Across Their Entire Organization for Lasting Impact
Bucureşti, Bucureşti, România
11 K urmăritori
Peste 500 de contacte
11 K urmăritori
Peste 500 de contacte
Servicii
Articole de Emil Muthu
Activitate
11 K urmăritori
-
Emil Muthu a publicat aceastaOn Google, you target a keyword. On Meta, you target an audience. In ChatGPT, the system can understand what someone is actually trying to do from the conversation they are having right now. If someone is comparing software, planning a purchase, researching a business problem or evaluating different options, the ad system can use that context to decide what is relevant in that moment. And when personalized ads are enabled, that decision can also use selected signals from previous ChatGPT interactions. That is a different model from buying a placement. You are not saying: "Put my ad next to this keyword." You are giving the system context about where your product might be useful, and the platform decides when the conversation creates the right moment. OpenAI is already moving beyond basic CPM and CPC buying into conversion optimization and automated bidding. The platform is also scaling fast. ChatGPT Ads reached a $1B annualized revenue run rate in less than 200 days. More than 1 billion people use ChatGPT every week. https://lnkd.in/dgPehb8w And self-service advertising is now expanding across India, Europe, the Middle East and North Africa. OpenAI itself still describes the platform as early and says it is continuing to learn and improve it. So a new advertising model is being built while it already operates at real scale. The question for marketers will soon be different. Not: "What keywords should we bid on?" But: "In which conversations should our company genuinely be useful?"
-
Emil Muthu a distribuit aceasta„Romania is still mostly consuming AI rather than creating it. We are becoming users of other people’s technology instead of building our own, and that is a serious missed opportunity”, says Alexandra Cernian, Associate Professor at Politehnica University of Bucharest and European Commission project evaluator, in the latest REWIRED episode. Without that shift in mindset, our local tech ecosystem risks remaining just a consumer market for global solutions, rather than a key player generating high-value innovation. In this episode of REWIRED, Alexandra Cernian, Ph.D, Habil. breaks down the gap between just adopting AI and actually creating it. We have the talent and the technical expertise. The real question is: what is holding us back from making the leap from consumers to creators?
-
Emil Muthu a distribuit aceastaAI has changed the economics of software. https://lnkd.in/d9qtZdYS Until recently, software cost was predictable. You had 100 employees, you paid for 100 licenses, you roughly knew what next month would cost. Now you can wake up to thousands of euros in extra spend without buying anything new. We had this happen recently. Anthropic changed which model the default setting resolves to in part of our Claude Code setup, over to Opus 5. Our code-review tooling kept doing exactly what it did before. We didn't add a feature. We didn't start a new project. Nobody clicked upgrade. The cost difference hit hundreds of euros a day. That's the reality every leader needs to sit with. AI isn't a subscription anymore. It's variable, automated consumption, and sometimes it moves because of a decision made outside your company. The paradox is the models keep getting cheaper. Your bill can still climb if you send every task to the most powerful model, leave automations on default, and don't know what each process costs. The fix isn't less AI. It's not sending everything to the most expensive model. The strong model plans, makes the hard calls, checks the final output. The cheap model classifies, extracts, reformats, runs the repetitive volume. Pin the model. Track the cost per process. When usage spikes, someone gets a flag. So the question isn't "what does the license cost" anymore. It's: which of my processes can spend money on their own? What can a vendor change tomorrow without my approval? And what work am I paying frontier prices for that doesn't need it? That's where the margin gets lost, or won back.Microsoft reports are exposing AI's real cost problem: Using the tech is more expensive than paying human employees | FortuneMicrosoft reports are exposing AI's real cost problem: Using the tech is more expensive than paying human employees | Fortune
-
Emil Muthu a distribuit aceastaI've seen too many AI initiatives start with the technology and end without a business case. Alexandra Cernian, Ph.D, Habil. puts it better than I could in this REWIRED episode. If your team is debating AI adoption, start here. https://lnkd.in/dYU_3C9nEmil Muthu a distribuit aceasta"AI is not a business strategy. And if a company cannot clearly define the business problem AI is meant to solve, that company may not be ready to adopt it. " That’s how Alexandra Cernian, Ph.D, Habil. , Associate Professor at Politehnica University of Bucharest and long-time European Commission evaluator, frames it in the new REWIRED episode: a mistake she keeps seeing in AI adoption. Because AI adoption should not start with: “Where can we use AI?” It should start with: “What business problem are we trying to solve?” Without that answer, an AI strategy risks becoming just another technology initiative without a clear business outcome. In the new episode, Alexandra will join REWIRED to discuss what companies are getting wrong about AI adoption. New episode of REWIRED with Alexandra Cernian is live now. Link in the comments. ### Neurony is a technology company with over 20 years of experience working with clients worldwide, and in recent years, we have increasingly provided AI-based services. Through the REWIRED podcast, we aim to explain the impact Artificial Intelligence has on our lives, whether we are talking about running a business, technology, consumer behavior, education, leadership, or how we spend our free time.
-
Emil Muthu a distribuit aceastaAI adoption is not the same thing as business value. HCLTech asked 500 enterprise decision-makers. 90% see AI changing how work gets done. 18% report a real dent in revenue. https://lnkd.in/dv7hRiau A second study, 639 AI leaders this time, found the same gap from a different angle. In 57% of companies, returns still aren't outrunning what's being spent on AI. https://lnkd.in/duhPgQEy For me, that's actually good news. Because the gap doesn't seem to be about which model someone bought. The companies getting results almost always have two things: a use case you can measure, and someone senior who actually owns it. So for every AI project, three questions. What number should move, who owns that number, by when. Not "productivity." Conversion rate. Error rate. Quoting time. Margin. Cost per transaction. No baseline, no owner, no date, and what you have isn't a business investment. It's an experiment. Leadership's job isn't to approve more AI. It's to tie it to an outcome you can point to on a P&L.
-
Emil Muthu a distribuit aceastaIn the latest episode of REWIRED, the Neurony podcast, Marius Ciurariu and I spoke with Miruna-Cristina BujorBujor, Engagement Manager at McKinsey&Company, about what makes AI projects successful. Miruna points out that there are several critical challenges: having a clear, business-led strategy, scaling AI smartly and integrating it with core systems, making sure you have access to high-quality data. But the most underestimated challenge is often the human side: getting people on board early, making sure they understand their roles, providing the right upskilling. And most importantly: addressing the fear of the unknown. 🎙️ Watch the latest REWIRED episode with Miruna Bujor, Marius Ciurariu, and me to hear the full conversation: https://lnkd.in/dtfpmsVk ### Neurony is a technology company with over 20 years of experience working with clients worldwide, and in recent years, we have increasingly provided AI-based services. Through the REWIRED podcast, we aim to explain the impact Artificial Intelligence has on our lives, whether we are talking about running a business, technology, consumer behavior, education, leadership, or how we spend our free time. #REWIRED #AI #ArtificialIntelligence #AITransformation #Leadership #Neurony
-
Emil Muthu a distribuit aceastaOne of the biggest potential mistakes companies make when adopting AI is that people use it in a very fragmented way, for painful but not significant enough tasks, without truly transforming the way they work. Details about how things should be done are discussed in the most recent REWIRED episode, in which our guest was Miruna-Cristina Bujor, Engagement Manager at McKinsey & Company. Link to the new episode in the comments.
-
Emil Muthu a distribuit aceastaMy colleagues created a very nice carousel about how we work with clients on Artificial Intelligence implementation. A carousel about how we make this process work: from prototype to business results. Read through the slides and, if you have the time, the text on our blog as well: https://lnkd.in/dbD-s5iq If you don’t have any time at all, here’s a summary: 1. Define success before building anything 2. Start with one focused workflow 3. Give every AI system a clear owner 4. Check readiness before you build 5. Prioritisation requires the right people in the room.
-
Emil Muthu a distribuit aceastaMiruna-Cristina Bujor, Engagement Manager at McKinsey & Company, thinks that the single most understated challenge in any AI rollout is people. I think the same. And many of the guests featured in the 20 episodes of our REWIRED podcast so far believe the same. Precisely because my colleagues and I believe so strongly in preparing people for Artificial Intelligence, last year we launched the REWIRED podcast and our AI-in-business training programs. And I hope that, on Thursday, August 13, at 4:00 PM, many of you will watch our podcast episode featuring Miruna Bujor from McKinsey & Company, one of the world’s leading business consulting firms, because you’ll learn a lot about the use of AI in business.Emil Muthu a distribuit aceasta"The single most understated challenge in any AI rollout is not integration or data quality, it is people: buy-in, upskilling, and fear of the unknown role that follows". This is one of the key ideas from the latest episode of the REWIRED podcast, which we are releasing on Thursday, August 13, featuring Miruna-Cristina Bujor, Engagement Manager at McKinsey & Company. Miruna has ten years of experience working with banks and insurance companies across Romania, CEE, and Western Europe, and since last year, her work has focused on helping organizations in these sectors adopt Artificial Intelligence. We’ll learn more about the projects she has been involved in and what she has learned from them on Thursday, at 4:00 PM, from Miruna-Cristina Bujor and our hosts, Marius Ciurariu and Emil Muthu. 👉Register below to be notified when this new episode of REWIRED îs live. ### Neurony is a technology company with over 20 years of experience working with clients worldwide, and in recent years, we have increasingly provided AI-based services. Through the REWIRED podcast, we aim to explain the impact Artificial Intelligence has on our lives, whether we are talking about running a business, technology, consumer behavior, education, leadership, or how we spend our free time.
-
Emil Muthu a apreciat aceastaEmil Muthu a apreciat aceastaA good podcast is one where you forget about the cameras 🎥 We’ve talked about the problems with adopting AI in an organisation, especially in Romania where only 5 out of 100 businesses use AI (compared to the EU average of 20) Why is that? Well, a lack of structure in our own thinking and processes PLUS the wishful thinking that AI will magically figure it out by itself. But we also took the time to discuss larger themes, even landed on the 17th century mathematics of Newton and Leibniz which are the fundamental backbone of how AI models are trained today. Thanks Marius Ciurariu and Emil Muthu for this discussion 🤘
-
Emil Muthu a apreciat aceastaEmil Muthu a apreciat aceastaPalantir built an entire business on one promise: when your system makes a consequential decision, you can prove why. That capability ships as an enterprise contract, on someone else's infrastructure. Someone rebuilt that layer in Python and put it on PyPI under MIT. It's called Semantica. Point it at your data — files, Postgres, Kafka, email, Git, Databricks, Snowflake — and it extracts entities and relationships into a knowledge graph instead of a pile of embeddings. Every agent decision becomes a first-class node: record it, link it to its causes, then trace the full causal chain back months later. Every fact carries W3C PROV-O provenance, so the audit trail exports in a format compliance teams already accept. Graph construction, reasoning, and provenance are fully deterministic — no LLM required anywhere in that path. Storage is polyglot and swappable without touching your code: Neo4j, FalkorDB, Apache AGE, AWS Neptune, or RDF via Oxigraph, Jena, Blazegraph. Ships with an MCP server, a REST API, a CLI, and a browser-based graph explorer.
-
Emil Muthu a apreciat aceastaEmil Muthu a apreciat aceastaPrivate pool or resort pool? Games room or cinema room? Exact Home or Allocation on Arrival? At the British Airways Holidays Florida Trade Market in Newcastle, we asked travel agents a series of quick-fire questions about the choices that shape an Orlando villa recommendation. Welcome to #WeAskedTheTrade: a new Ocean Beds series bringing the voices of the travel trade into conversations about villa product, distribution and customer choice. Over the coming weeks, we’ll unpack their answers, explore the real considerations travel agents navigate when matching the right villa product to the right customer, and share the Ocean Beds perspective behind each choice. Which question should we explore first? Thank you so much to every agent who decided to take part - it was a pleasure meeting you all 💙 #WeAskedTheTrade #OceanBeds #VillaDistribution #TravelTrade #OrlandoVillas