Iván Martínez

Privacy is a top concern when discussing ChatGPT-like tools with professionals. With this working proof of concept, I show how open-source models and tools like @langchain enable a 100% local execution, ensuring your data never leaves your environment. github.com/imartinez/priv…

🚀 MAJOR NEWS: Strategic partnership with @TelefonicaTech to revolutionize enterprise AI! When one of Europe's largest tech giants validates your approach, you know the market is shifting toward private AI workspaces where data sovereignty is ensured 🔐 telefonicatech.com/en/news/we-enh…

🫵 Join us for our next event on #PrivateAI hosted by @GoogleStartups Madrid on June 9th at 17:00 for great talks and even greater food and drinks networking afterwards 🍻 Register now! lu.ma/7sym9e4r

Understanding MCP architectures helps teams make smarter AI integration decisions. The key factor is where the LLM processes data. This diagram different setups - no "wrong" choice if you're informed about the tradeoffs. More details: zylon.ai/blog/mcp-archi…

@CarlesRamio @espublico Muy buen artículo @CarlesRamio. La única alternativa real es "Private AI": una AI que no utiliza la nube, sino que está 100% circunscrita al entorno seguro de la administración. Es nuestra misión en @ZylonPrivateGPT, y trabajamos en una exploración inicial con la administración.

Los peligros de la utilización del chat GPT por parte de los empleados públicos Carles Ramióadministracionpublica.com/los-peligros-d… a través de @espublico

Episode 1 of our "Zylon Success Stories" series with @UCBerkeley's @GoldmanSchool who built ChatGSPP using our open-source PrivateGPT tech - a fully private #AI chatbot running on-prem that helps students & faculty access information in a quick & easy way youtu.be/2X3TvcVWAnA

@Mascobot @a16z Would love to test @ZylonPrivateGPT on that beast! A match made in heaven. SSH access and a couple hours is all it'd take to bring the world's most powerful on-prem AI platform into your setup. Seriously, drop me a dm if you're up to it!

🚨 New: We @a16z built an 8x RTX 4090 GPU AI workstation from scratch —compatible with the new RTX 5090 with PCIe 5.0, for training, deploying, and running AI models locally— so you don’t have to. Here’s how we built it, why it matters, and how you can build one too. Full guide👇

🤯 What our team at @ZylonPrivateGPT just built has shattered every boundary of on-prem enterprise AI. Agentic workflows, real-time collaboration, editable canvas, inline AI support, multi-modality... ... 100% on YOUR servers. No more "quality or privacy" dichotomy. 💪

💪 Yesterday: "I need to do this." 🦾 Today: "I need to ask AI to do this." 🧠 Tomorrow: "I need to explain the problem to AI so it can solve it." 🤖 Near future: "I need to update AI on our company vision so it can identify and solve our next critical problems." 🗒️👇

With us your company data NEVER leaves your servers, no matter what LLM we use. Our on-prem AI runs ENTIRELY local in YOUR infrastructure with zero outbound connections. It's not marketing—it's architecture. True data sovereignty means complete control. zylon.ai/blog/why-aren-…

Announcing @MistralAI Small 3.1: multimodal, multilingual, Apache 2.0, the best model in its weight class. 💻 Lightweight: Runs on a single RTX 4090 or a Mac with 32GB RAM, perfect for on-device applications. 🗣️ Fast-Response Conversations: Ideal for virtual assistants and other applications where quick, accurate responses are essential. ⚡ Low-Latency Function Calling: Capable of rapid function execution within automated or agentic workflows. 🔍 Specialized Fine-Tuning: Customizable for specific domains. 🧠 Advanced Reasoning Foundation: Inspires community innovation, with models like DeepHermes 24B by Nous Research built on Mistral Small 3. We're releasing base and instruct checkpoints for Mistral Small 3.1 to support further downstream customization of the model.

Announcing the latest DeepHermes Preview models, DeepHermes 24B and 3B! huggingface.co/collections/No… These new models are Hybrid Reasoners - meaning you can toggle ON and OFF the long chain of thought reasoning whenever you want a short, intuitive answer, or a long, well reasoned higher accuracy answer, now available on our API and to download on HuggingFace.

🔍 Knowing beats guessing every time. We've made significant strides with Opik by @Cometml. Our latest iteration brought a 52% improvement in context recall and enhanced measurement capabilities. Iterating, measuring, and optimizing—this is how we drive excellence to #PrivateAI

Just launched: RFPs automation with Zylon's Bulk Q&A Agent Flow! 🚀 Our new private AI feature extracts questions from RFPs, searches your knowledge base, and generates accurate responses with sources—all in minutes. See how engineering firms are saving 1000+ hours annually:

Today, we release QwQ-32B, our new reasoning model with only 32 billion parameters that rivals cutting-edge reasoning model, e.g., DeepSeek-R1. Blog: qwenlm.github.io/blog/qwq-32b HF: huggingface.co/Qwen/QwQ-32B ModelScope: modelscope.cn/models/Qwen/Qw… Demo: huggingface.co/spaces/Qwen/Qw… Qwen Chat: chat.qwen.ai This time, we investigate recipes for scaling RL and have achieved some impressive results based on our Qwen2.5-32B. We find that RL training con continuously improve the performance especially in math and coding, and we observe that the continous scaling of RL can help a medium-size model achieve competitieve performance against gigantic MoE model. Feel free to chat with our new models and provide us feedback!

@gregisenberg Fully agree. Companies today choose the SaaS that most closely fit their workflows, and adapt to it. In a matter of 1-2y, no one will adapt their workflows to external software, but build the in-house "software" that boosts those workflows. The key piece will be the Platform.

SaaS is being dismantled as we speak! We're witnessing the slow-motion collapse of an entire business model that dominated tech for two decades. The $1.3 trillion SaaS is being quietly hollowed out from within by AI agents. Here's how I see it playing out: Phase 1 (Now): AI as co-pilot. We're seeing this everywhere, Copilot for developers, Gamma for presentations, Harvey for legal research etc. These AI layers sit atop existing software, making it more efficient. The SaaS companies feel safe, even excited, as AI seems to make their products more valuable. They're bringing knives to what they think is a knife fight. Phase 2 (Next 12-18 months): The agent invasion. AI moves from co-pilot to autonomous operator. They're replacement workers that can fully operate existing software on your behalf. The dam breaks when someone can say "analyze our Q2 performance" rather than clicking through Tableau, or "optimize our ad campaigns" instead of navigating Meta's ad manager. The expertise previously bundled with the software gets unbundled by agents. Phase 3 (2-3 years): Software invisibility. The final phase happens when the agents bypass the human interfaces altogether. Why render dashboards, buttons and menus when AI can just access the APIs directly? The value proposition of SaaS, bundling software, workflow, and expertise into user-friendly interfaces unravels completely. The interfaces were designed for humans, but agents don't need them. Most SaaS incumbents don't see it coming because this isn't a classic disruption pattern. It's not about competing products with better features. It's about the evaporation of the core assumption that humans will operate software. What's more, the barrier to creating custom, internal software is collapsing simultaneously. Companies that once had to choose between expensive custom development or off-the-shelf SaaS can now spin up bespoke solutions in days instead of months. Why pay Hubspot $1,500/month for a CRM when your team can build 'HubspotForUs' with an AI coding assistant over a weekend? The same features, perfectly tailored to your workflow, with no ongoing subscription costs. This democratization of software creation means every company becomes a potential software producer rather than just a consumer. The specialized knowledge that SaaS companies monopolized is now available to anyone with access to an AI coding agent and domain expertise. It went from $1M to build an MVP to build a SaaS to basically free overnight. I bet the metrics will be puzzling at first, DAUs remain strong while feature usage mysteriously declines. The power users who drive revenue suddenly need fewer seats. Customer success calls shift from "how do I use this feature?" to "can your software work with my AI agent?" Or worse: "we built our own version that better fits our workflow." The survivors won't be those with the best features or even those who add AI features fastest (from no AI to "ai-assisted"). The winners will be companies that expose their software's capabilities through agent-friendly APIs and position themselves as the most trustworthy information sources and execution engines in their domain. There's also the shift from monthly subscriptions to outcome based software (pay per outcome, pay per task etc) but that's a tweet for another day! The $1T question: Will Microsoft, Atlassian, Adobe etc. successfully navigate this transition, or will they be the Digital Equipment Corporation of our era too invested in the previous paradigm to adapt to the new one? All I know is this will be a golden era for startups in the space. SaaS is being dismantled, piece by piece, workflow by workflow, interface by interface. Am I wrong?

8.5% of AI prompts contain sensitive data. Yet, 64% of employees use Cloud AI tools like ChatGPT without safeguards. CISOs, ignoring this = risk ⚠️ Stop compromising your most valuable assets by using Cloud-based AI providers. #PrivateAI is the solution to secure your data 🔐

Using our on‑prem Private AI through @ZylonPrivateGPT, I can confidently connect to our knowledge base and ask questions—ensuring that all input, inference, and output remain securely within our infrastructure. 🧵

@cocktailpeanut @lmstudio Very interesting concept, this is where AI UX should be heading. Everyday users shouldn't be deciding whether or not to enable "reasoning" for each question. Intent and complexity detection are key. Thanks for sharing!

DeeperHermes DeepHermes can act as a simple AI agent or a deep thinking AI, just by passing a system prompt. But what if we can go DEEPER? The LLM decides on its own how to respond by...asking itself! (No system prompts) I made a gradio app powered by @lmstudio, it works!

DeepHermes 3 Preview is exactly what I was hoping for! A single model that switches between "normal" and "reasoning" modes via system prompt, built on Llama 3.1 8B. More efficient than loading separate models—crucial for #PrivateAI. huggingface.co/NousResearch/D…