Nikolaos Vasiloglou

An important article I wrote that differentiates LLM assistance and agentic AI. Please read. 🔗 https://lnkd.in/eb8mJyjJ A note for those applying to jobs: This matters for your career too. When you're in an interview and someone asks what you built, be precise. If you used an LLM to help you write code faster, say that, don't call it an agent. If you built a RAG pipeline, call it retrieval-augmented generation, don't call it agentic. If you built a chatbot with a system prompt, own it for what it is. Hiring managers who know the difference will respect your precision. Hiring managers who don't know the difference need you to teach them. Either way, you win by naming it correctly. Overselling what you built doesn't just hurt your credibility when someone digs deeper, it tells the room you don't actually understand the architecture you're describing. Know what you built. Know what it's called. Know why it matters at that level. That's what gets you hired. #AI #AgenticAI #LLM #NLP #SportsAnalytics #DataScience #CareerAdvice #MachineLearning

13

Over the holiday break, I spent some time revisiting some of the foundational work shaping the field of scientific AI. One paper that stands out—published back in 2021—was co-authored by Damien Guého, now a Senior R&D Engineer at Geminus. The paper tackles a timeless challenge: how do you uncover the true physics of a system when all you have is sparse, noisy, imperfect data? Damien and his co-author's approach is both clever and elegant. Instead of a brute force deep learning strategy, they leveraged clever math to uncover a physics-grounded structure with interpretable dynamics—from messy, nonlinear signals. The result is a robust, reliable and generalizable solution that uses a fraction of the computational effort. They first solved the problem using a deep learning approach that required 7,720 parameters. They then applied their method that used just 3 terms and was 10,000× more accurate. This mindset is core to Geminus' mission to simplify engineering and operational decision making in complex systems where uncertainty abounds and data is sparse and messy. Work like that of Damien and his colleagues continues to inspire where we’re headed next. Proud of the team at Geminus Labs—and excited for what’s coming. Check out his paper here. https://lnkd.in/eHt-5tXX Watch this space - be sure to follow Geminus to be part of the conversation.

32

AI terms simplified - we in the industry come up with ridiculously complex terms to keep the club exclusive but you would recognize most of these words if presented simply: RAG - Answer via searching through documents Embeddings / Vector DBMS - Representation of information in numbers for easier computation Context Engineering - Systems architecture to provide the right information to LLM Memory Layer - Database frameworks with low latency and search built in AI Agents - Autonomous workflows where decisions are taken or guided by Large Language Models Its important we explain these terms simply so enterprises and users of all shapes and sizes can envision how it would help them solve their problems. Only then businesses can feel AI's benefit.

2

Another breakthrough in #AI was published earlier today by Sapient Intelligence, and the linked headline by VentureBeat says it all. https://lnkd.in/giYJtEb9 TL;DR, the new #HierarchicalReasoningModel (#HRM) reasons more like a human by quickly thinking about the best way to solve a problem and then thinking more slowly to solve the problem. This new method can be trained on an order of magnitude fewer training examples and can solve multi-step puzzles that simply befuddle #CoT #LLM s, like #ChatGPT and #Claude, ultimately requiring far fewer resources.

6

2 Comments

97% cheaper, 20x faster, with guaranteed business outcomes. Tech services and AI coding tools are in for a rude awakening when it comes to pricing, because we’re going to shatter their margins and pass nearly all of it on to our customers. A sneak peak under the hood.

11

This year at 𝐓𝐞𝐜𝐡𝐧𝐢𝐜𝐚 𝟮𝟬𝟮𝟱, we set out to build AI that adapts to people. We created 𝗥𝗢𝗠𝗔-𝗩𝗟𝗠, a multimodal, memory-augmented AI agent featuring 𝗿𝗲𝗰𝘂𝗿𝘀𝗶𝘃𝗲 𝗿𝗲𝗮𝘀𝗼𝗻𝗶𝗻𝗴, adaptive speech + vision processing, Qdrant-powered agentic memory, and an accessibility-first interface designed for users with dyslexia and color-vision differences. We were grateful when a representative from Intuit visited our demo, appreciated our approach, and encouraged us to consider internship and full-time opportunities which was a meaningful reminder that accessible AI has real impact beyond the hackathon. Proud of what we built, and excited to keep creating AI that’s smarter, more adaptive, and truly built for everyone. Couldn’t have done this without the amazing team - Sarvesh Khetan, Manoj Kumar Bashaboina and Rupali Patel. (Github link in comments!) #AI #Accessibility #MachineLearning #TechForGood #AgenticAI #VLM #DSPy #InclusiveTech #Technica2025 #MLH #WomenInTech #DeepLearning

54

7 Comments

Voice AI is quietly hitting an inflection point 🔊 NVIDIA just published a great deep dive on scaling speech recognition for real-world voice agents - latency, streaming accuracy, robustness, and why “ASR quality” isn’t just about WER anymore. 👉 https://lnkd.in/enDUKHvG A few things that stood out: - Streaming + low latency is non-negotiable if you want natural conversations - ASR needs to handle interruptions, background noise, accents, and partial utterances, not clean demo audio - Scaling voice agents isn’t just model quality, it’s systems engineering If you want to actually try building or running voice agents instead of just reading about them, Pipecat is a great place to start: 👉 https://pipecat.ai This is the kind of progress that makes voice agents finally feel… practical. Not demos. Not prototypes. Production. Curious to see where the ecosystem lands this year 👀

149

10 Comments

Last year, I posted about 𝗶𝗺𝗮𝗴𝗲 𝘀𝗰𝗮𝗹𝗶𝗻𝗴 𝗽𝗿𝗼𝗺𝗽𝘁-𝗶𝗻𝗷𝗲𝗰𝘁𝗶𝗼𝗻 𝗮𝘁𝘁𝗮𝗰𝗸𝘀 on multi-modal AI Agents, displaying texts in images that are only visible to the AI (not humans) during their down-sampling process. Now you can try it yourself, as someone made generation of such images more easily accessible. I am all for people to play around with it responsibly for awareness of the risks of using multi-modal browser agents. Just be careful not to be logged into sensitive places when using a browser agent on platforms where an image prompt-injection could likely occur. https://lnkd.in/d-RgqsAm

7

2 Comments

Last week, Rasa and Nebius Academy co-hosted a webinar, and nearly 100 engineers attended to discuss AI agents and Claude Code. Thanks to Denis Volkhonskiy and Stan Fedotov, PhD, for the deep dive into Claude Code. I walked away with sharper mental models myself. Here is what stuck with me: 1. An agent is not just an LLM Brains (LLM) plus Knowledge (RAG) plus Tools (Bash, file I/O) plus Workflow control. All four layers have to work together before you have something production-ready. 2. Plan Mode changes everything Stop asking agents to "fix the bug." Enter a collaborative planning state first. The agent proposes a path, you approve it, and then it executes. No surprises. At Rasa, we call this equivalent Conversation-Driven Development. Listen before you build. The same principle applies to agents. 3. Security is not an afterthought When you give an agent a Bash tool, you're handing it the keys to your house. Two non-negotiables: never expose .env files, and vet every MCP skill like you'd vet a random script from the internet. 4. Context is your budget. Spend it deliberately. Use “compact” or “clear” between sub-tasks. Tag specific files with “@” instead of pointing at the whole repo. "Look at everything" equals hallucinations plus a bill you didn't expect. What's next We are headed into a future of digital coworkers, conversational and headless. The engineers who figure out how to combine both are going to have an unfair advantage. We're building a hyper-local community of AI engineers in London. If you're in London, the link is in the comments. Plus, don't forget to apply to the AI Performance Engineering program from the Nebius Academy, the deadline is approaching!

44

2 Comments