A few things we’ve written down.

AI Is Rewriting Textbooks. But Are We Also Rewriting How We Think? Last week, Google launched Learn Your Way an experiment to rethink one of education’s most overused yet under-loved tools: the textbook. 𝐀𝐧𝐝 𝐡𝐞𝐫𝐞’𝐬 𝐭𝐡𝐞 𝐩𝐫𝐨𝐛𝐥𝐞𝐦 𝐢𝐭’𝐬 𝐭𝐫𝐲𝐢𝐧𝐠 𝐭𝐨...

Full Stack of a RAG System Retrieval-Augmented Generation (RAG) is one of the most practical ways to make large language models more reliable. But behind the scenes, a full RAG system has many moving parts. Think of it as an iceberg the frontend is visible, but most of the...

Today, I’m sharing some bittersweet news: I’ve decided to move on from my role as 𝐏𝐫𝐢𝐧𝐜𝐢𝐩𝐚𝐥 𝐀𝐮𝐭𝐨𝐦𝐚𝐭𝐢𝐨𝐧 𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫 at Ally Bank. My time here has been transformative. Working at the intersection of 𝐀𝐠𝐞𝐧𝐭𝐢𝐜 𝐀𝐈, 𝐈𝐧𝐭𝐞𝐥𝐥𝐢𝐠𝐞𝐧𝐭...

Agentic AI vs AI Agents, What's the Real Difference? Let’s break it down in system terms: 𝐀𝐈 𝐀𝐠𝐞𝐧𝐭𝐬: These are rule-based or goal-directed components that perform single, predefined tasks using hardcoded tools and static workflows. ➜ No planning ➜ No autonomy ➜ No memory...

Chunking is the unsung hero of RAG systems. Everyone talks about retrieval. Few talk about how you prepare the data you’re retrieving. But here’s the thing: If your LLM gives vague, irrelevant, or hallucinated answers—it’s often not the model’s fault. It’s the chunking. Let’s...

At some point, LLMs stop just getting better - and start getting weirdly smarter. This idea - called emergent abilities - is not just AI hype. It is one of the most fascinating shifts happening as we scale language models. Researchers from Google and Stanford recently dug into...

LangChain just dropped something big: Open Deep Research an autonomous agent built for long-form, reliable research. This is not just a wrapper around web search. It is a modular, multi-agent system that scopes, investigates, and synthesizes knowledge with real structure....

Every powerful AI agent is built on a pattern. And no, I do not mean code. I mean design thinking. 𝐀𝐬 𝐚𝐠𝐞𝐧𝐭𝐬 𝐠𝐞𝐭 𝐦𝐨𝐫𝐞 𝐜𝐚𝐩𝐚𝐛𝐥𝐞, 𝐭𝐡𝐞 𝐫𝐞𝐚𝐥 𝐪𝐮𝐞𝐬𝐭𝐢𝐨𝐧 𝐢𝐬 𝐧𝐨𝐭: “𝐂𝐚𝐧 𝐢𝐭 𝐝𝐨 𝐭𝐡𝐞 𝐭𝐚𝐬𝐤?” 𝐈𝐭’𝐬: “𝐂𝐚𝐧 𝐰𝐞 𝐚𝐫𝐜𝐡𝐢𝐭𝐞𝐜𝐭 𝐢𝐭...

AI agents don’t just predict. They act. And that changes the security game entirely. We’re entering a new era Where AI agents aren’t just responding to prompts, They’re planning, reasoning, and executing real-world tasks. And that means: New power. New risks. New rules. Google’s...