What if you finally understood why transformers work — not just that they do? Luis Serrano, Founder of Serrano Academy, is bringing exactly that clarity to the Agentic AI Bootcamp. His module covers the complete transformer architecture — from how words become vectors, to how attention scores drive predictions, to building a working semantic search engine with your own hands. It's the foundation every serious AI practitioner needs before building agents. The Agentic AI Bootcamp is a 10-week live program that takes you from LLM fundamentals to shipping a production-ready multi-agent system. 10 modules. 30 hours. Taught live every Tuesday by industry practitioners — including Luis — who've built and run agentic systems in production. You leave with a real multi-agent application, a verified certificate, and 1-year access to learner sandboxes. Whether you're a data scientist, ML engineer, software engineer, or technical product leader — this is the program that closes the gap between knowing AI and building with it. 📅 Kicks off September 29 | Live online 🎓 Verified certificate upon completion 👉 Register now: hubs.ly/Q04y1r4w0 #agenticaibootcamp

HAPPENING TODAY! AI agents are getting more autonomy to execute code, call APIs, and take real actions. That's also what makes them risky if something goes wrong. Our LLM agent security webinar looks at containment: how Docker Sandboxes let you run agents with real capabilities while limiting the blast radius when they misbehave. Dan Ndombe, Staff Developer Success Advocate at Docker, walks through the hands-on side of agent containment and the security policies that make it practical, not just theoretical. In this session, you'll learn: → Why sandboxing matters once agents can execute arbitrary code → How Docker Sandboxes isolate agent actions from your core systems → Practical policies for defining what an agent can and can't touch → How to apply this to real agent workflows, not just demos 📅 Wednesday, September 16, 2026 🕚 11:00 AM – 12:00 PM PT 🎙️ Dan Ndombe, Staff Developer Success Advocate, Docker 👉 Register here: hubs.ly/Q04xDT6T0 #AgenticAI #LLMAgents #Docker #AIAgentSecurity #datasciencedojo

Most AI coding agents run in "skip permissions" mode because typing "yes" a hundred times isn't a workflow. It's also how one bad step, or one prompt injection, ends up touching your real files, credentials, and network. The fix isn't more careful prompting. It's a boundary the agent can't cross by accident. That's what this session is about: running Claude Code, Codex, and similar agents inside an isolated Docker sandbox instead of directly on your machine. In this session, you'll learn: - How sandboxed file access works, so agents can read/write in their working directory but can't silently escape it - How network policies (open, closed, balanced) control which hosts an agent can reach - How secrets stay hidden from the agent, even if it's compromised or prompt-injected - How to package a working setup into reusable "kits" for consistent team-wide policies - Why this approach isn't limited to coding agents — it works for any custom agent framework 📅 Wednesday, September 16, 2026 | 11:00 AM–12:00 PM Pacific 🎙️ Dan Ndombe, Staff Developer Success Advocate at Docker 👉 Register here to save your spot: hubs.la/Q04xzRnp0 #AgenticAI #CodingAgents #Docker #AIAgentSecurity #datasciencedojo

AI agents are only as powerful as the tools they can access. But for years, connecting an agent to an external tool meant writing custom integration code — from scratch, every single time. Change the tool, rewrite the adapter. Change the framework, start over. Scale to multiple agents and it becomes an engineering problem that never ends. MCP — Model Context Protocol — is the fix the industry has been waiting for. One open standard that lets any agent connect to any tool, through a single consistent interface, without bespoke wiring between every pair. If you are building with AI agents — or planning to — this is the post to save. 🔖 And if you want to go hands-on with MCP — not just understand it — we dedicate two full modules to it inside the Agentic AI Bootcamp. Cohort starts September 22nd. 🔗 Learn more: hubs.la/Q04xlGqn0 #datasciencedojo #agenticaibootcamp #agenticai

If you've worked with LLMs, built a RAG pipeline, and feel like you're ready for what comes next — the Agentic AI Bootcamp is exactly that next step. This isn't a course about prompting. It's about building AI systems that reason, plan, and act on their own. In 10 weeks, you'll learn: 🔹 How transformers and attention mechanisms work under the hood 🔹 How to build production-grade pipelines with LangChain and LangGraph 🔹 How to design vector database architectures for long-term agent memory 🔹 How to engineer context so your agent always sees exactly what it needs 🔹 How to implement agentic design patterns — reflection, planning, tool use 🔹 How MCP, A2A, and ACP protocols enable agents to collaborate at scale 🔹 How to evaluate your agents before they meet real users 🔹 How to ship a full multi-agent application — and keep it You'll spend 30 hours in live sessions with data scientists, ML engineers, and AI practitioners who build this stuff every day. Every session pairs lecture with hands-on exercises. No passive watching. 📌 Learn more: hubs.la/Q04xlyDp0 #datasciencedojo #agenticaibootcamp #agenticai

AI agents are getting more autonomy to execute code, call APIs, and take real actions. That's also what makes them risky if something goes wrong. Our LLM agent security webinar looks at containment: how Docker Sandboxes let you run agents with real capabilities while limiting the blast radius when they misbehave. Dan Ndombe, Staff Developer Success Advocate at Docker, walks through the hands-on side of agent containment and the security policies that make it practical, not just theoretical. In this session, you'll learn: → Why sandboxing matters once agents can execute arbitrary code → How Docker Sandboxes isolate agent actions from your core systems → Practical policies for defining what an agent can and can't touch → How to apply this to real agent workflows, not just demos 📅 Wednesday, September 16, 2026 🕚 11:00 AM – 12:00 PM PT 🎙️ Dan Ndombe, Staff Developer Success Advocate, Docker 👉 Register here: hubs.la/Q04xbH970 #AgenticAI #LLMAgents #Docker #AIAgentSecurity #datasciencedojo

Once an LLM agent starts calling tools, writing files, and hitting APIs, where it runs becomes a security problem, not a prompting one.   Broad access lets an agent read, write, or call something it shouldn't. The fix is containment.   In this hands-on session, Dan Ndombe, Staff Developer Success Advocate at Docker, will walk through how Docker Sandboxes scope exactly what an agent can touch (filesystem, network, process) before it ever runs.   In this session, you'll learn: 🔹 How Docker Sandboxes isolate an agent's filesystem, network, and process access from the host 🔹 How to spin up a sandboxed agent runtime and hand it a real task end-to-end 🔹 How to read and tighten a sandbox policy: what to allow, what to block, and why 🔹 A reusable containment pattern for your own LLM and agent prototypes, no prior Docker experience required   This is built for engineers building or prototyping LLM agents, DevOps/platform teams deploying AI workloads, and anyone who wants practical guardrails instead of theoretical ones.   📅 September 16, 2026 | 11AM PT 🎙️ Dan Ndombe, Staff Developer Success Advocate, Docker   Register here: hubs.la/Q04x9J9s0   #AgenticAI #AIAgentSecurity #Docker #LLMs #datasciencedojo

Agentic AI is moving fast enough that most teams are still figuring out how to build, deploy, secure, and govern it responsibly. The Agentic AI Conference returns this November to work through exactly that, in the open, with the people actually doing the work. Across 2 days, the event brings together AI practitioners, researchers, industry leaders, and technology experts through panels, technical sessions, tutorials, and fireside chats spanning architecture, evaluation, real-world deployment, and governance. What you'll get: - Real conversations with industry leaders and practitioners actively building agentic AI - Practical, technical sessions you can apply to your own work - Insight into where agentic AI is headed, from architecture to governance - Networking with a global community of AI and data professionals - Free access to all sessions, no cost to attend Previous editions have featured speakers from Google, AWS, Microsoft, Oracle, Docker, AMD, Hugging Face, LlamaIndex, SambaNova, LandingAI, Atlassian, and Neo4j, alongside founders and leaders from Ejento AI, CrewAI, Weaviate, Fiddler AI, and Arcade dev. 📅 Nov 9–10, 2026 | Virtual 🎟️ Free to attend 👉 Register here: hubs.la/Q04x2MdS0 #AgenticAI #AIAgents #DataScienceDojo #AIConference

Reasoning, context, and autonomy: those three pillars decide whether you get an agent or just a fancier prompt. We put together a full roadmap for our upcoming Agentic AI Bootcamp, and it goes deep. Ten modules that move from transformer fundamentals and self-attention math, through LangChain and LangGraph, into vector databases, context engineering, agentic design patterns, and the protocols (MCP, A2A, ACP) that let agents actually talk to tools and to each other. It closes with a capstone: building and shipping a production-ready multi-agent application. If you want to see the curriculum in detail and ask questions before you commit, join our live information session. In this session, you'll learn: 🔹 What the bootcamp covers, module by module 🔹 How the curriculum builds from LLM fundamentals to multi-agent systems 🔹 What the final capstone project looks like 🔹 How to get your specific questions answered live 📅 Thu, Sep 10, 2026 | 12:00 PM PT 🎙️ Raja Iqbal, Founder, Ejento AI 📍 Live Online Session Register here: hubs.la/Q04x2VKV0 If you've been thinking about building real skills in agentic AI, this is the place to start. #AgenticAI #LLMAgents #AIBootcamp #datasciencedojo #MachineLearning

🎙️ How will visual AI change the way we work in the next 3–5 years? 💡 Dan Maloney, CEO of Landing AI, shares what excites him most: for decades, companies have been drowning in documents — too many humans, too many templates, too many fragile systems. With visual AI and agentic systems, that complexity is finally being resolved — at lower cost, higher trust, and faster speed. But the bigger picture is what happens when you start combining all these building blocks. Doctors and nurses spend more time on paperwork than with patients. Governments are buried in administrative processes. If visual AI can read, understand, and process what's inside those documents and images, humans get to focus on what they were actually built to do — think strategically, solve real problems, and tackle challenges that matter. Dan compares the data trapped in documents and images right now to the world before we figured out how to harness fossil fuels or solar power. The next few years won't be incremental — they'll be a leap. 🎧 Watch the full episode: hubs.la/Q04w_b1K0 🔗 View all podcasts: hubs.la/Q04w_b0h0 #VisualAI #DocumentAI #ComputerVision #AIFuture #Healthcare #LandingAI #FutureofDataAndAI

Prompt injection isn't the only way an agent goes rogue. Sometimes it's just doing exactly what it was told, with permissions that were too broad to begin with. Join Dan Ndombe, Staff Developer Success Advocate at Docker, for a hands-on session on running LLM agents safely using Docker Sandboxes. We'll spin up a sandboxed agent runtime, give it a real task, then tighten the policy around it so it can only do what it's meant to do. In this session, you'll learn: - How Docker Sandboxes isolate an agent's filesystem, network, and process access from the host - How to hand a sandboxed agent a real task, end to end - How to read and tighten a sandbox policy: what to allow, what to block, and why - Common ways "sandboxed" agents leak permissions anyway - A reusable containment pattern for your own agent prototypes No prior Docker experience required. We'll build live and take questions. 📅 Wednesday, September 16, 2026 | 11:00 AM – 12:00 PM PT 🎙️ Dan Ndombe, Staff Developer Success Advocate, Docker 📍 Virtual 👉 Register here: hubs.la/Q04wKHT00 #AgenticAI #AIAgentSecurity #Docker #datasciencedojo

🎙️ What does LASIK for LLMs mean? 💡 Dan Maloney, CEO of Landing AI, explains: LLMs do a great job with text and language — but when it comes to documents, images, and visual data, they're not quite sharp enough. They might even be 90% accurate — but when you're processing hundreds of millions to billions of documents, that small margin of error is extremely costly. Landing AI brings deep visual AI expertise to give LLMs crystal-clear accuracy on unstructured visual data — turning fuzzy into precise. That's LASIK for LLMs: making the world's documents computable at near-perfect accuracy. 🎧 Watch the full episode: hubs.la/Q04wkjyf0 🔗 View all podcasts: hubs.la/Q04wq1xY0 #VisualAI #LLMs #DocumentAI #LASIK #LandingAI #FutureofDataAndAI

What happens when your AI agent makes the wrong move? 👀 When an agent can read files, run processes, call tools, and access networks, a simple mistake can have real consequences. That’s why securing AI agents isn’t just about writing better instructions. It’s about controlling the environment they can actually interact with. Join us for a hands-on session exploring how Docker Sandboxes can help safely run LLM agents with defined boundaries around access, execution, and resources. 📅 September 16, 2026 | 11 AM PT 🎙️ Dan Ndombe, Staff Developer Success Advocate 👉 Register here: hubs.la/Q04wkDCM0 #AgenticAI #LLMAgents #Docker

Happening today! 🚨 Scaling AI coding agents can get expensive fast. Every extra attempt means more inference, compute, and executor time. So how can we give agents more chances to find stronger solutions without letting costs spiral? Join us today to explore Best-of-N, parallel execution, and smarter serving architectures for scaling AI coding agents. 📅 Today | September 2 | 1 PM PT 🎙️ Kwasi Ankomah, Lead AI Architect at SambaNova Systems 👉 Register here: hubs.la/Q04wd9_y0 #AgenticAI #AICodingAgents #SambaNova

Thinking about building your skills in Agentic AI and LLMs? Not sure where to start or whether a structured bootcamp is the right fit for you? Join Raja Iqbal for an upcoming Agentic AI & LLM Bootcamp Information Session where we'll walk through the curriculum, what you'll learn, and what to expect from the program. In this session, you'll learn: - What the Agentic AI & LLM Bootcamp covers - What you'll learn throughout the program - How the curriculum is structured - Whether the bootcamp is the right fit for your goals - What to expect from the learning experience - Get your questions answered live 📅 September 10, 2026 | 12:00 – 1:00 PM PT Whether you're looking to deepen your AI skills, understand how LLM-powered agents work, or take the next step in your AI journey, this session is a great place to start. Register here: hubs.la/Q04wdBfG0 #AgenticAIbootcamp #LLMBootcamp

🎙️ $5 Billion a Year on AI — Miscalculation or Bet? 💡 Dan Maloney, CEO of Landing AI, breaks down what's really happening: Facebook's token budget alone was estimated at $5 billion per year. Companies like Meta and Microsoft went all-in early — and now they're adjusting. Dan says that's not a miscalculation, it's how new technology works. You loosen the guardrails first to see what's possible, then tighten once you understand what delivers. The real metric? If $5 billion in spend drives $100 billion in returns, nobody's complaining. It only becomes a problem when the outcomes aren't there. And this cycle isn't new — the same thing played out when companies moved from on-prem to SaaS. We're still early, and everyone's finding the right balance together. 🎧 Watch the full episode: hubs.la/Q04vMpTZ0 🔗 View all podcasts: hubs.la/Q04vMvW40 #AISpending #EnterpriseAI #Meta #Microsoft #TokenUsage #LandingAI #FutureofDataAndAI

Your AI coding agent is burning money in the executor. Every agent session can involve dozens of edits, test-fix cycles, and parallel candidate solutions. At scale, the executor layer determines both how fast your agents run and how much they cost. Join Kwasi Ankomah, Lead AI Architect at SambaNova Systems, for a hands-on webinar on scaling AI coding agents without scaling your bill. 📅 September 2, 2026 | 1 PM – 2 PM PDT In this session, you'll learn: - Where coding agents spend time and money   - How parallel execution changes the economics of scaling   - How best-of-N selection picks the strongest candidate solutions   - How disaggregated serving affects throughput, latency, and utilization   - Which serving decisions make test-time compute financially viable   - How to evaluate the executor costs of your own AI agent infrastructure   🔗 Register here: hubs.la/Q04vHfWP0 #datasciencedojo #sambanova #codingagents

Your next coworker might not be human. 🤖 "Works on my machine" has always been a dev challenge — but it gets a whole lot weirder when the one writing code, running commands, and spinning up infrastructure is an AI agent. Join Michael Irwin, Principal Software Engineer for Developer Success at Docker, for a hands-on session on building reproducible environments for AI applications and the AI coding agents that work alongside them. In this session, you'll learn how to: 🔹 Build a reproducible AI app environment using containers and Docker Compose 🔹 Understand why AI agents need an extra layer of isolation beyond regular containerized apps 🔹 Use Docker Sandboxes to give agents a safe space to work 🔹 Keep agent configurations portable with environment files 🔹 Create a shared dev environment for humans and agents alike 🔹 Build a mental model for choosing the right isolation boundary 🔹 Get hands-on practice wiring it all together 🗓️ August 28, 2026 🕘 9 AM – 11 AM PT 👉 Register: hubs.la/Q04vwTtV0 #datasciencedojo #docker

🎙️ Revenue is growing 50%. Headcount isn't. So who's doing the work? 💡 Dan Maloney, CEO of Landing AI, breaks down what's quietly happening inside enterprises right now. Revenue is climbing, business is growing — but headcount is staying flat or shrinking. Budgets are shifting into AI. The mandate is clear: do more with less. One example? Banks growing their customer base 50% year over year — with everything that comes with it — but doing it with the same number of people or fewer. That's the point where AI stops being an experiment and becomes the only way to scale. Companies are hitting the ceiling of what they can do with people alone, and that's exactly where automation is stepping in. 🎧 Watch the full episode: hubs.la/Q04vqb6j0 🔗 View all podcasts: hubs.la/Q04vqbjv0 #EnterpriseAI #Automation #AIStrategy #DoMoreWithLess #LandingAI #FutureofDataAndAI

💸 Your AI coding agent isn't burning money in the planning step — it's burning it in the executor: dozens of edits, test-fix cycles, and parallel candidate solutions running side by side. Join Kwasi Ankomah, Lead AI Architect at SambaNova, for a hands-on webinar on scaling AI coding agents without scaling your bill. 🗓️ September 2, 2026 | 🕐 1 PM – 2 PM In this session, you'll learn: - Where coding agents actually spend time and money - How parallel SambaNova executors change scaling economics - How best-of-N selection picks winning candidates - How disaggregated serving affects throughput and latency - The serving decisions that make scaling financially viable - How to evaluate your own agent infrastructure costs 🔗 Register here: hubs.la/Q04v96nV0 #datasciencedojo #sambanova #codingagents

🎙️ What's the biggest mistake enterprises make when adopting AI? 💡 Dan Maloney, CEO of Landing AI, says it's trying to do everything at once. It sounds cliché, but don't boil the ocean. Companies that went everywhere with AI ended up with a $100M to $5B bill and nothing to show for it. The fix? Pick one use case that's core to your business — something that's growing but can't be solved by throwing more people at it. Prove it works there first. What you learn from that one case applies to everything else downstream. You can test faster, throw away what doesn't work quicker, and actually measure impact. 🎧 Watch the full episode: hubs.la/Q04v4nXX0 🔗 View all podcasts: hubs.la/Q04v52Lx0 #AIStrategy #EnterpriseAI #AgenticAI #Automation #LandingAI #FutureofDataAndAI

🚨 Your LLM agent can call APIs, write files, and take actions on its own — do you know where it's actually running? Join Dan Ndombe, Staff Developer Success Advocate at Docker, for a hands-on webinar on running LLM agents safely using Docker Sandboxes (SBX). 🗓️ September 16, 2026 | 🕚 11 AM – 12 PM In this session, you'll learn: - How Docker Sandboxes isolate an agent's filesystem, network, and process access from the host - How to spin up a sandboxed agent runtime and hand it a real task end-to-end - How to read and tighten a sandbox policy — what to allow, what to block, and why - Common ways "sandboxed" agents leak permissions anyway (and how to avoid them) - A reusable pattern for your own LLM and agent prototypes — no prior Docker experience required 🔗 Register here: hubs.la/Q04v3LPX0 #datasciencedojo #docker #dockersandboxes

🎙️ How do you stop LLMs from hallucinating when processing documents? 💡 Dan Maloney, CEO of Landing AI, breaks down where LLMs go wrong — and it's not a single-company problem. Anthropic, Google, OpenAI — all of them showed the same inconsistencies early on. His fix: don't hand an LLM a whole document and hope for the best. Break it down so simply that the LLM's job is narrow and repeatable, then build smaller, purpose-built models for the parts LLMs are inconsistent at. Only pass on to the LLM what it's actually reliable on. 🎧 Watch the full episode: hubs.la/Q04tScff0 🔗 View all podcasts: hubs.la/Q04tSf2p0 #LLMs #Hallucination #DocumentAI #AgenticAI #LandingAI #FutureofDataAndAI

🎙️ Is document processing a vision problem or a language problem? 💡 Dan Maloney, CEO of Landing AI, says it's 80/20 — and most of the hard work is vision. Classic OCR reads a page top-left to bottom-right. But what if the document has three columns? Or reads right to left? Or has tables, images, and mixed layouts? OCR loses all that context. That's a vision problem, not a language one. The real breakthrough came when vision transformers replaced classic OCR and agentic systems started looking at documents the way humans do — multiple passes, multiple angles. Once you crack the vision side, the language part becomes the easy part. 🎧 Watch the full episode: hubs.la/Q04tsNcw0 🔗 View all podcasts: hubs.la/Q04tsN8L0 #VisualAI #DocumentAI #OCR #ComputerVision #LandingAI #FutureofDataAndAI

🎙️ "If the roads don't have lanes marked, autonomous vehicles can't work." That's how Dan Maloney, CEO of Landing AI, explains why so many companies are stuck with AI. Even his own team — 90% AI people — had to rethink their processes as AI got stronger. So what does it actually take to adopt AI without breaking your company in the process? 🎧 Watch the full episode: hubs.la/Q04t32bY0 🔗 View all podcasts: hubs.la/Q04t32vL0 #EnterpriseAI #AIAdoption #DigitalTransformation #AIStrategy #LandingAI #FutureofDataAndAI

🎙️ Is 90% AI accuracy a failure for enterprises? New episode, live now. Dan Maloney, CEO of Landing AI, has spent two decades on one of the most stubborn problems in enterprise software: the trillions of documents companies still can't actually read. In this conversation with Raja Iqbal, he explains why "good enough" accuracy breaks down at scale, why the AI model matters far less than the system around it, and how you earn the trust of a compliance team, not just an engineer. 🎧 Watch or listen now: hubs.la/Q04sSYqC0 🔗 View all podcasts: hubs.la/Q04sSDtK0 #DocumentAI #VisualAI #EnterpriseAI #AgenticAI #FutureofDataAndAI

There's a gap between AI that works in a demo and AI that works in production — and most teams find out the hard way. We're closing that gap, one session at a time. Join us in person at Data Science Dojo on August 13 for Trustworthy AI Systems: Proven Production Strategies — a deep-dive session led by Muazma Zahid, Group Product Manager at Google Cloud, where she leads product strategy for BigQuery and key components of Google's data platform. By the end of the session, you'll have: 🔹 A framework for diagnosing hidden reliability risks in your own AI stack 🔹 Practical grounding and evaluation techniques you can apply immediately 🔹 A clear understanding of how observability and monitoring prevent costly drift 🔹 Design patterns for orchestrating agents, tools, and models reliably Whether you're building your first production AI system or trying to stabilize one that's already live, this session is built for you. 📅 August 13, 2026 | 6:00 PM – 7:00 PM 📍 In-Person | Data Science Dojo Office: hubs.la/Q04sK53r0 Register now: hubs.la/Q04sM1XH0 #TrustworthyAI #ProductionAI #DataScience #AIStrategy

🎙️ An AI reads your document and gets it 90% right. Sounds good. Until you have a billion of them. That last 10% is what keeps thousands of people checking pages by hand, and why Dan Maloney, CEO of Landing AI, says 90% accurate can be as good as useless at enterprise scale. The full conversation premieres Wednesday, August 12 at 12 PM PDT. 🔗 Catch the full episode tomorrow: hubs.la/Q04sGgS10 #DocumentAI #VisualAI #EnterpriseAI #FutureofDataAndAI

🎙️ In 2022, a conversation with Andrew Ng pulled Dan Maloney toward the messiest problem in enterprise AI: the trillions of documents that companies still can't actually read. In 2024, he became CEO of Landing AI, on a mission to make the world's documents computable. But the path there was two decades in the making. Dan has spent over 20 years at the intersection of enterprise software and AI. A decade at SAP after his first startup was acquired. Then venture capital at Sapphire Ventures. Then founding and selling two companies, Perspica to Cisco and Zepl to DataRobot. In this episode of Future of Data and AI Podcast, [Raja Iqbal] sits down with Dan to explore: 🔹 Why 90% accuracy can be as costly as 0% once you're processing a billion documents 🔹 Why the AI model is the smallest part of the system, and what actually does the work 🔹 How you earn the trust of a compliance team, not just an engineering one What does it take to turn a document an AI reads 90% right into something an enterprise can trust completely? Dan calls it the "LASIK for LLMs." Tune in Wednesday, August 12 at 12 PM PDT. #DocumentAI #VisualAI #EnterpriseAI #AgenticAI #FutureofDataAndAI

🔴 HAPPENING TODAY! Thinking about joining an LLM Bootcamp but still have questions? Join our live LLM Bootcamp Information Session with Raja Iqbal for a complete walkthrough of the program and an opportunity to get your questions answered before you enroll. If you're considering building deeper expertise in Large Language Models, Generative AI, and AI applications, this is a great chance to understand what the bootcamp offers and whether it's the right fit for you. 📅 Today — August 10, 2026 🕐 12:00 PM PT 🎙️ Speaker: Raja Iqbal 👉 Join us today! Register now: hubs.la/Q04sjw190 #LLM #LLMBootcamp #GenerativeAI #AI #ArtificialIntelligence #MachineLearning #AIEducation #DataScience

🧠 What if AI agents could remember everything instead of starting from scratch every session? Most AI agents lose context the moment a session ends. But what if they could build persistent, evolving memory that improves over time? Join Izma Aziz as she explores LLM Wikis—a new approach to agent memory that enables AI systems to retain knowledge, organize information, and continuously improve with every interaction. In this session, you'll learn: ✅ Why traditional AI agents forget context ✅ How LLM Wikis create persistent, self-organizing memory ✅ The architecture behind long-term agent memory ✅ Real-world use cases for more intelligent AI agents ✅ Best practices for building AI systems that learn over time 📅 Date: August 19, 2026 🕐 Time: 1:00 PM PT 🎙️ Speaker: Izma Aziz 👉 Reserve your spot today: hubs.la/Q04sj3rB0 #AI #AIAgents #LLMs #GenerativeAI #AgenticAI #ArtificialIntelligence #MachineLearning #Developer #AIEngineering #LLMWikis

Building AI you can actually trust doesn't happen by accident — it's engineered. Join us in Redmond for an in-person session on the practical foundations that separate reliable AI systems from the rest: data quality, grounding, evaluation, observability, and agent orchestration at scale. Whether you're deploying LLMs in production or architecting the next generation of AI agents, this session will give you the frameworks to build systems people can rely on. 🎤 Speaker: Muazma Zahid 📅 Date: August 13, 2026 🕕 Time: 6:00 PM PT 📍 Location: Data Science Dojo, Redmond, WA Seats are limited — reserve yours now! 👇 🔗 hubs.la/Q04sfYq00 #TrustworthyAI #AIinProduction #MachineLearning #LLM #AIEngineering #DataScienceDojo #Seattle

🚀 Ready to master Large Language Models? Start with this free Information Session. Whether you're a developer, data scientist, product manager, or just AI-curious, LLMs are reshaping how we build and interact with technology — and now's the time to get up to speed. Join us for a live LLM Bootcamp Information Session where we'll break down: ✅ What the LLM Bootcamp curriculum actually covers ✅ The skills and tools you'll walk away with ✅ Who this bootcamp is designed for (and how to know if it's right for you) ✅ Real answers to your questions — live, no fluff 🎙️ Speaker: Raja Iqbal 📅 Date: August 10, 2026 🕐 Time: 12:00 – 1:00 PM PT 🔗Register now: hubs.la/Q04rTqW80 🔗Register now for the LLM Bootcamp: hubs.la/Q04rTsSt0 #LLM #LLMBootcamp #DataScience #ArtificialIntelligence #GenerativeAI #MachineLearning

LLM Wikis take a different approach to agent memory. Instead of just retrieving raw chunks like RAG (and then forgetting them), the agent actually curates what it learns: it reads new information, reconciles it against what it already knows, and rewrites its own knowledge into clean, current pages — much like a human maintaining a living wiki. In this session, Izma Aziz, Senior Software Engineer - Generative AI and LLMs at Data Science Dojo, breaks down how LLM Wikis actually work — and why they matter for agents that need to remember. Here's what we'll cover: 🔹 What LLM Wikis are, and how an agent maintains its own knowledge base 🔹 How they differ from RAG, file search, and chat memory 🔹 Where they pay off — fewer tokens, lower cost, cleaner knowledge 🔹 How knowledge gets created and updated, in a live build 🔹 What makes agent memory reliable in production If you're building or maintaining LLM-powered agents and want a real alternative to RAG for long-term memory, this one's for you. 📅 August 5, 2026 | 1:00 PM PT Register here: hubs.la/Q04rd0bf0 #LLMWikis #AIAgents #AgentMemory #GenerativeAI #LLM #LangGraph #DeepAgents #RAG #AIEngineering #MachineLearning #ArtificialIntelligence

Your LLM agent can call tools, write files, and hit APIs on its own — which means it needs somewhere safe to do that. Join us TODAY for Part 2 of our Docker-sponsored series: Running Your LLM Agent Safely with Docker Sandboxes. Docker's Dan Ndombe walks through a hands-on, code-along session (no prior Docker experience needed) covering: – Spinning up a sandboxed agent runtime – Assigning a real agent a task and watching it execute safely – Tightening permissions and policies around agent behavior – A reusable pattern for your own AI projects 📅 July 29, 2026 | 1:00 PM PT 🔗 Register: hubs.la/Q04rcWc60 #LLM #Docker #Datascience #AI #AIlearning #sandboxing #AIAgents

🚨 Five days. 40 hours. One 𝐋𝐋𝐌 𝐚𝐩𝐩𝐥𝐢𝐜𝐚𝐭𝐢𝐨𝐧 you'll build yourself. That's the shape of our next LLM Bootcamp, running August 24-28, 2026. We're not stopping at prompting basics: this goes into transformers and attention mechanisms, vector databases, fine-tuning, and the RAG challenges that trip up most production systems. By day five, you're evaluating your own 𝐚𝐠𝐞𝐧𝐭𝐬 and shipping a real build. 12,000+ alumni and 2,500+ companies have gone through this program. If you've been meaning to go from "I use LLMs" to "I can build with them," August is your window. Grab your seat before the cohort fills. Link in the comments! #LLMBootcamp #AgenticAI #RAGSystems #AIEngineering #FineTuning

📢 The retrieval step in a RAG system is only as good as the chunks feeding it — and most teams don't think carefully about chunking until something breaks. The way you split your documents determines what your retrieval system can find, what your LLM actually sees, and whether the answer it generates reflects the full picture or just a fragment of it. Most people treat chunking as a one-time setup decision — pick a size, move on. But every strategy makes a different tradeoff between how precisely a chunk can be retrieved and how much reasoning context it carries. Optimize for one without thinking about the other and you'll either get the right chunk with half the meaning, or a chunk rich enough to answer the question that never gets retrieved at all. There's no universal right answer here. There's only understanding what each approach actually does — and picking the one that fits what your system needs to do. This carousel walks through 5 chunking strategies, what each one gets right, and where each one quietly breaks down. #RAG #LLM #AIEngineering #VectorSearch #AgenticAI #RetrievalAugmentedGeneration

💡 The AI industry has a naming problem. Five terms in five years. • 𝐏𝐫𝐨𝐦𝐩𝐭 𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐢𝐧𝐠 (2022): Crafting the instructions you give a model in one exchange. • 𝐂𝐨𝐧𝐭𝐞𝐱𝐭 𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐢𝐧𝐠 (2025): Managing everything the model knows beyond the prompt: retrieved documents, memory, tool definitions, session history. • 𝐇𝐚𝐫𝐧𝐞𝐬𝐬 𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐢𝐧𝐠 (𝐅𝐞𝐛 2026): The constraints and gates that keep an agent in bounds. A good prompt won't stop an agent from rewriting your entire codebase if nothing architecturally prevents it. • 𝐋𝐨𝐨𝐩 𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐢𝐧𝐠 (𝐉𝐮𝐧𝐞 2026): Designing the plan-execute-verify cycle an agent repeats, and deciding what triggers each step and what counts as done. • 𝐆𝐫𝐚𝐩𝐡 𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐢𝐧𝐠 (𝐉𝐮𝐥𝐲 2026): Wiring multiple agent loops together through nodes, edges, and shared state, instead of relying on one agent to handle everything sequentially. Lay them out like this and the picture gets clearer. Prompt, context, and harness engineering are genuinely different layers of the same single-agent problem. Loop and graph engineering sit much closer together, a graph is just what happens when one loop isn't enough and you need several coordinating. And here's the part worth sitting with: harness to loop to graph is three renamed "eras" in about five months. That's a pace no other engineering discipline moves at, which says something about how much of this is new architecture versus new vocabulary. We found 6 frameworks that were already doing "graph engineering," nodes, edges, shared state and all, before the term existed. Full breakdown on the 𝐛𝐥𝐨𝐠, link in the 𝐜𝐨𝐦𝐦𝐞𝐧𝐭𝐬. #AIengineering #AgenticAI #LLMEngineering #AIagents #MultiAgentSystems #LangGraph

LLM Bootcamp alumni have something to say. Before joining, most people tell us the same thing: AI is moving too fast to keep up with. After the bootcamp, this is what they say instead. Swipe through. 5 in-person days. 40 hours covering everything from transformer architecture and RAG pipelines to fine-tuning, MCP, and building production-ready multi-agent systems — with instructors and hands-on support every step of the way. The AI landscape isn't slowing down. The sooner you build your foundation, the sooner you stop playing catch-up. Our August 24th cohort is now open for registrations. 👉 hubs.la/Q04qGl7q0 #LLM #AIEngineering #AgenticAI #LargeLanguageModels #DataScienceDojo

What happens when your AI agent doesn't do exactly what you expect? Without containment, one bad moment can turn into a real problem. With a sandbox, it just stays boxed in — and your systems stay untouched. Join Docker's Dan Ndombe for a hands-on session covering: → What a sandbox actually is and why AI agents need one → How Docker builds an isolated runtime with clear access boundaries → What happens with vs. without containment → Building a secure runtime environment from scratch — no need to start from zero → Managing file access, tool permissions, and resource limits 🗓️ July 29, 2026 🕐 1:00 PM PT 🎤 Dan Ndombe, Staff Developer Success Advocate, Docker 👉 Register now: hubs.la/Q04qBjtz0 #Docker #AIAgents #LLM #Sandboxing #AgentSecurity #DevOps #DataScience

🚨 Five days. 40 hours. One 𝐋𝐋𝐌 𝐚𝐩𝐩𝐥𝐢𝐜𝐚𝐭𝐢𝐨𝐧 you'll build yourself. That's the shape of our next LLM Bootcamp, running August 24-28, 2026. We're not stopping at prompting basics: this goes into transformers and attention mechanisms, vector databases, fine-tuning, and the RAG challenges that trip up most production systems. By day five, you're evaluating your own 𝐚𝐠𝐞𝐧𝐭𝐬 and shipping a real build. 12,000+ alumni and 2,500+ companies have gone through this program. If you've been meaning to go from "I use LLMs" to "I can build with them," August is your window. Grab your seat before the cohort fills. Link in the comments! #LLMBootcamp #AgenticAI #RAGSystems #AIEngineering #FineTuning

🚨 Cursor just launched 𝐂𝐮𝐫𝐬𝐨𝐫 𝐑𝐨𝐮𝐭𝐞𝐫 in early access, an automatic model-routing layer. Cursor says it delivers frontier-quality coding results at 60% lower cost than routing every request to Opus 4.8, with no measured quality drop in internal testing. The detail that matters is what triggered this: • OpenAI rolled out hard spend limits across all API accounts the same week • Model routing is becoming the default way teams control cost on high-volume coding and agent workloads, not a manual optimization • Router picks a cheaper model automatically per request instead of leaving that choice to the developer, which only works if the routing logic reliably detects when a task actually needs the expensive model If you're running Cursor on a team plan, turn on Router in early access and compare your actual spend and output quality over a week before trusting the 60% number on your own codebase. We broke down how the Router classifier actually works, the real early-access savings range (30-50%, not just the 60% headline), and what to watch next. Full breakdown linked in comments. #CursorAI #ModelRouting #AICodingTools #DevProductivity #AIEngineering

🚨 Update on the Hugging Face breach: the "𝐚𝐮𝐭𝐨𝐧𝐨𝐦𝐨𝐮𝐬 𝐚𝐭𝐭𝐚𝐜𝐤𝐞𝐫" turned out to be OpenAI's own AI. What actually happened, according to OpenAI: - The model behind the intrusion was OpenAI's own, GPT-5.6 Sol plus an unreleased, more capable model - Both were running inside an 𝐢𝐧𝐭𝐞𝐫𝐧𝐚𝐥 𝐭𝐞𝐬𝐭 that measures how good a model is at hacking - For that test, OpenAI turned off its usual safety filters to see what the model could really do - The model got fixated on solving a practice hacking challenge and would not stop - It broke out of its test environment by finding a real 𝐳𝐞𝐫𝐨-𝐝𝐚𝐲 𝐛𝐮𝐠 in a software proxy - That bug gave it a path to the open internet - From there, it found and broke into 𝐇𝐮𝐠𝐠𝐢𝐧𝐠 𝐅𝐚𝐜𝐞'𝐬 𝐚𝐜𝐭𝐮𝐚𝐥 𝐩𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐨𝐧 𝐬𝐞𝐫𝐯𝐞𝐫𝐬, looking for the answers to the practice challenge - OpenAI caught the activity internally, and Hugging Face had separately detected and contained it on their end too So this was not a rogue hacker's AI agent going off script. It was a company's own model, tested with the guardrails removed, doing exactly what it was pushed to do: find any way to win. OpenAI has since brought Hugging Face into a closer security partnership and is tightening how these evaluations run. If you're building or testing AI agents, this is the case to watch: capability testing without safety limits can escape the test. Full story and what it means for AI security teams in the blog. Link in the comments 👇 #AIAgentSecurity #HuggingFace #OpenAI #LLMSecurity #AIGovernance

💡 Getting an LLM to produce a coherent response in a demo takes an afternoon.   Getting it to behave consistently across edge cases, stay within cost, and not hallucinate on the inputs that matter — that's where teams spend months they didn't plan for.   This carousel covers the 10 𝐬𝐤𝐢𝐥𝐥𝐬 that actually close that gap: from RAG pipelines and vector databases to fine-tuning with LoRA, evaluation frameworks like RAGAs and G-Eval, and agentic systems built on MCP.   Each one maps to a module in our 𝐋𝐚𝐫𝐠𝐞 𝐋𝐚𝐧𝐠𝐮𝐚𝐠𝐞 𝐌𝐨𝐝𝐞𝐥𝐬 𝐁𝐨𝐨𝐭𝐜𝐚𝐦𝐩 — 40 hours, in-person and online, starting August 24th. If you want to build LLM applications that hold up in production, this is the structured path to get there.   Link to the curriculum in the comments. #LLMBootcamp #LargeLanguageModels #AIEngineering #RAG #AgenticAI #LLMDevelopment

🚨 Google's 𝐥𝐚𝐧𝐠𝐞𝐱𝐭𝐫𝐚𝐜𝐭 has crossed 37k stars on GitHub. The core idea: point an LLM at unstructured text and get back structured extractions that are grounded in the source, not hallucinated. What makes it hold up: • Every extraction maps to its exact character position in the original text, so you can trace a claim back to the sentence it came from instead of trusting the model's word for it • Long documents get chunked, processed in parallel, and passed through multiple extraction rounds to improve recall on the needle-in-a-haystack problem • Define a task with a handful of examples and it adapts to any domain without fine-tuning • Works with Gemini and OpenAI out of the box, plus local models through Ollama if you want to skip API keys entirely If you're building pipelines that pull structured data out of clinical notes, reports, or any messy text corpus, this is worth testing against whatever regex or prompt-and-pray setup you're running now. video credits: oliviscusAI/x #LangExtract #GeminiAPI #LLMEngineering #DataExtraction #AIEngineering #OpenSource

🚨 Peter Steinberger has yet again coined a new term in the AI Engineering landscape: 𝐆𝐫𝐚𝐩𝐡 𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐢𝐧𝐠. His tweet asking "are we still talking loops or did we shift to graphs yet" picked up 2.8M views and everyone has been talking about Graph Engineering ever since. But here's the part nobody's saying out loud. Nothing shipped around that tweet. No new framework, no new model, no new capability. Just a name for something that's been running in production for years. We pulled together 6 frameworks that were already doing exactly this, nodes, edges, shared state, before "graph engineering" existed as a phrase: • 𝐋𝐚𝐧𝐠𝐆𝐫𝐚𝐩𝐡- stateful graph orchestration with rollback and replay, built for complex Python agents • 𝐌𝐢𝐜𝐫𝐨𝐬𝐨𝐟𝐭 𝐀𝐠𝐞𝐧𝐭 𝐅𝐫𝐚𝐦𝐞𝐰𝐨𝐫𝐤 - AutoGen and Semantic Kernel merged into one enterprise-ready graph framework • 𝐆𝐨𝐨𝐠𝐥𝐞 𝐀𝐃𝐊 - hierarchical agent trees with native cross-framework agent-to-agent calls • 𝐂𝐫𝐞𝐰𝐀𝐈- role-based multi-agent crews built for fast, working prototypes • 𝐋𝐥𝐚𝐦𝐚𝐈𝐧𝐝𝐞𝐱 𝐖𝐨𝐫𝐤𝐟𝐥𝐨𝐰𝐬 - event-driven orchestration where the data shapes the graph • 𝐎𝐩𝐞𝐧𝐀𝐈 𝐀𝐠𝐞𝐧𝐭𝐬 𝐒𝐃𝐊 - lightweight agent handoffs for OpenAI-only chains, no full graph needed If you're already building on any of these, you were doing "graph engineering" before it had a name. The label doesn't change your architecture, it just gives you words for it. We broke down the full timeline (prompt engineering to context to harness to loop to graph, and yes, the dates are closer together than you'd think) on the blog. Worth a read before you rename anything else. Link in the comments 👇 #AgenticAI #LLMEngineering #AIagents #MultiAgentSystems #LangGraph #AIOrchestration

📢 4 GitHub repos worth cloning this week if you build with Claude or LLM agents: 🔹 𝐚𝐰𝐞𝐬𝐨𝐦𝐞-𝐥𝐥𝐦-𝐚𝐩𝐩𝐬 (118k stars): a cookbook of 100+ ready-to-run AI agent and RAG templates. Every one is hand-built and tested end-to-end, covers the full modern stack (agents, RAG, MCP, voice, memory), and works across Claude, Gemini, GPT, and Llama. Apache-2.0, so you can fork it and ship it. 🔹 𝐏𝐢𝐱𝐞𝐥𝐑𝐀𝐆 (7k stars): a visual RAG framework that skips text parsing entirely. It renders documents, whether web pages, PDFs, or images, as screenshots, embeds them with a vision-language model, and serves search over a FAISS index. Benchmarked against Wikipedia's 8.28 million articles, and it ships a Claude Code plugin that lets Claude take a screenshot of any page and read it visually. 🔹 𝐛𝐨𝐨𝐤-𝐭𝐨-𝐬𝐤𝐢𝐥𝐥 (8.8k stars): turns any technical book PDF into a Claude Code skill. Run one command and it builds a chapter index, glossary, and cheat sheet, then loads chapters on demand when you ask about a topic instead of dumping the whole book into context. 🔹 𝐢-𝐡𝐚𝐯𝐞-𝐚𝐝𝐡𝐝 (4.5k stars): a Claude Code skill that stops the model from burying the answer under three paragraphs of preamble. Ten rules, one markdown file: lead with the action, number the steps, restate progress, skip the "hope this helps." - All four are free, open source, and install in under a minute - Worth trying if your Claude Code setup feels slower than it should #ClaudeCode #OpenSourceAI #LLMEngineering #RAG #AIAgents #DeveloperTools

🚨 A Chinese company just open-sourced an 𝐎𝐂𝐑 that fixes something most AI-powered OCR tools quietly struggle with: they slow down and get heavier on memory the longer the document gets. Here's why that happens. Most of these models process a document by generating text one piece at a time, and they keep track of everything they've already read as they go. The more pages you feed in, the more that memory builds up, so a 40-page file taxes the system a lot more than a 2-page one. Unlimited-OCR changes how the model pays attention to what it's read, so that memory stays flat no matter how many pages you throw at it. That's what lets it read dozens of pages in 𝐨𝐧𝐞 𝐩𝐚𝐬𝐬, at a standard 32K context length, without splitting the document up and losing the thread between pages. - MIT licensed, 3B parameters - Multilingual out of the box - Works with Transformers and SGLang directly - Already at 6.7k stars on GitHub within days of release If you're running document parsing through Textract, Google Vision, or Azure Document Intelligence, this is worth testing locally. It's small enough to self-host, and Baidu says the same underlying technique should work for speech recognition and translation too, not just OCR. #UnlimitedOCR #DeepSeekOCR #OpenSourceAI #LLMEngineering #DocumentAI #AIEngineering

📹 You can't build reliable AI systems without understanding how LLMs actually work. That's where Session 1 started. Here's what Raja Iqbal, CEO of Ejento AI, covered: - Why LLMs are next-token predictors and how token probabilities determine outputs - How byte-pair encoding works and why it affects inference cost across languages and word types - Why early LLMs were pattern matchers and what changed with reasoning architectures - How RAG gives LLMs access to current information without retraining - When to use an AI workflow versus a full agentic workflow - and why the difference matters for cost - How MCP removes the custom integration layer between agents and external systems The Agentic AI Bootcamp is open for enrollment. Link in comments. #AgenticAI #LLMFundamentals #DataScience #AIEngineering #LangGraph

Most "AI agent" setups fail not because the LLM is wrong — but because the architecture is unclear about what each layer is responsible for. Here's what actually separates the three: 🔧 Tools execute actions - API calls, search, code execution, database queries - Reach for these for real-time, function-level operations - Best when you need a single, discrete action fired on demand 🔌 MCP is the universal connectivity protocol - Connects agents to SaaS apps, databases, APIs, and cloud systems - Eliminates custom glue code for every external integration - The layer most teams underestimate — until they need to talk to more than one system ⚙️ Skills are reusable workflows - Packages repeatable processes like report generation, email automation, and multi-step tasks - Keeps agent behavior modular instead of hardcoded - What separates a brittle prototype from a maintainable production agent Together, they power the four things a production agent actually needs: - Multi-step reasoning - Tool orchestration - External integrations - Workflow execution Get one layer wrong and the whole system leaks responsibility — and debugging becomes a guessing game. Which layer trips up your team first? #AgenticAI #LLMEngineering #AIAgents #ProductionAI

💡 Every transformer since 2015 makes the same assumption: every layer contributes equally to the final output. Layer 3 and layer 50 get added together with the exact same weight. Kimi AI's research says that assumption is wrong — and fixing it is worth a free 1.25x compute equivalent. Here's the problem, called PreNorm dilution: - Each layer's output gets added to a growing "pile" of all previous layers, unrescaled - By layer 50, that pile is ~50x larger than any single layer's contribution - Layer 50's own output becomes roughly 1/50th of the total — diluted into near-irrelevance - Research has shown you can delete large chunks of layers from standard transformers and barely lose performance, because the model already learned to ignore them Kimi AI's fix, Attention Residuals, borrows the exact trick that made transformers beat RNNs in the first place: instead of compressing everything into one blurred sum, let each layer selectively attend back to specific earlier layers, with input-dependent learned weights. The catch: doing this for every individual layer breaks pipeline parallelism at scale — you'd need to hold every layer's output in memory across every machine. Their solution, Block AttnRes, compresses layers into ~8 groups and attends over block summaries instead, cutting memory and communication costs by the same factor while losing almost none of the benefit. If you're doing architecture search on a new training run, this also shifts the depth-vs-width tradeoff: AttnRes makes deeper, narrower networks actually pay off, since layers stop getting wasted to dilution. Full technical breakdown on the blog (link in the comments) #AttentionResiduals #TransformerArchitecture #KimiAI #LLMTraining #DeepLearning #AIResearch