🎥 AI Agents and MCP¶
Welcome to the AI Agents and MCP section of the V2 Video Hub. Explore curated high-density videos with architectural summaries.
Table of Contents¶
- Red Hat OpenShift
- Generative AI and Large Language Models
- Grafana Assistant Claude Code Grafana MCP Server
- Red Hat OpenShift AI
- Claude Code
- Neural Networks
- LLM Architecture and Post-Training
- Agentic DevOps
Red Hat OpenShift¶
🎬 Thursday morning general session - May 9 - Red Hat Summit 2019
Architectural Summary
This session outlines the architectural enablement of cloud-native AI/ML workloads by integrating Red Hat OpenShift with NVIDIA GPU acceleration and automated MLOps platforms like H2O.ai and ProphetStor. It demonstrates how standardizing on a Kubernetes-based hybrid cloud substrate abstracts heterogeneous hardware environments, facilitating deterministic scaling, resource orchestration, and cognitive monitoring for high-performance AI pipelines. This unified operational model serves as a foundational blueprint for modern enterprise AI-platform engineering and edge computing architectures.
Generative AI and Large Language Models¶
🎬 Artificial Intelligence | 60 Minutes Full Episodes
Architectural Summary
This documentary anthology traces the rapid evolution of artificial intelligence from early deep learning implementations to advanced generative AI systems like Google's Bard and OpenAI's ChatGPT. For 2026 cloud-native architectures, these developments highlight the critical need for integrating scalable AI model orchestration, strict ethical guardrails, and secure data pipelines directly into enterprise platform engineering. Understanding these socio-technical shifts assists cloud architects in designing resilient, compliant AI-integrated infrastructures that balance massive computational demands with robust operational governance.
Grafana Assistant Claude Code Grafana MCP Server¶
🎬 I Stopped Staring at Dashboards. AI Reads My Grafana Metrics Now.
Architectural Summary
This video explores how AI agents can read Grafana metrics, logs, and traces directly using Grafana Assistant and Claude Code wired up to the Grafana MCP server. It demonstrates querying Prometheus, Loki, and Tempo, generating custom dashboards from natural language, and analyzing runtime data from the command line. This workflow bridges the gap between analysis and remediation, allowing agents to verify fixes and make data-grounded decisions.
Red Hat OpenShift AI¶
🎬 Red Hat OpenShift AI overview
Architectural Summary
Red Hat OpenShift AI provides an enterprise-grade MLOps platform built on Kubernetes that standardizes the training, tuning, serving, and monitoring of foundation and predictive AI models across hybrid and multi-cloud environments. By integrating open-source frameworks like Jupyter, PyTorch, and KServe with certified hardware accelerators, it delivers a secure, consistent, and self-service environment for platform and data science teams. This architecture ensures robust AI governance, operational scalability, and accelerated time-to-market for intelligent cloud-native applications.
Claude Code¶
🎬 Mastering Claude Code in 30 minutes
Architectural Summary
This technical session explores the architecture and implementation of Claude Code, Anthropic's agentic CLI designed for autonomous, multi-step engineering tasks. It details how the tool leverages the Model Context Protocol (MCP) to integrate with external data sources and documentation, enabling a self-healing development cycle where the agent autonomously reads code, executes shell commands, and iterates through test-driven development (TDD) loops. For 2026 platform engineers, mastering these agentic workflows is critical for scaling complex refactoring, managing cognitive load in large-scale repositories, and establishing robust human-in-the-loop (HITL) governance for AI-driven infrastructure operations.
Neural Networks¶
🎬 ¿Qué es una Red Neuronal? Parte 1 : La Neurona | DotCSV [SPANISH CONTENT]
Architectural Summary
This video deconstructs the foundational mathematical and algorithmic mechanics of a single artificial neuron, illustrating its direct relationship with linear regression, weights, biases, and activation functions. In a 2026 cloud-native landscape, mastering these core neural principles is critical for platform architects optimizing distributed micro-models and real-time AI inference engines deployed on Kubernetes-driven edge and cloud infrastructure. This granular understanding enables more efficient hardware acceleration profiling (GPUs/vGPUs/TPUs) and smarter resource allocation for decentralized machine learning pipelines.
LLM Architecture and Post-Training¶
🎬 Stanford CS229: Building Large Language Models (LLMs)
Architectural Summary
This Stanford CS229 technical deep-dive deconstructs the transition from raw autoregressive language models to instruction-tuned assistants, focusing on the systems orchestration required for 2026 AI infrastructure. It explores critical patterns in tokenization (BPE/Sub-word), parameter-efficient fine-tuning (PEFT/LoRA), and the shift from RLHF to Direct Preference Optimization (DPO) to simplify model alignment pipelines. For cloud architects, the lecture provides a foundational framework for optimizing the "Compute-to-Token" ratio and managing memory constraints (KV Cache) in distributed distributed inference environments, while advocating for LLM-as-a-Judge automated evaluation loops for scalable model governance.
Agentic DevOps¶
🎬 Agentic DevOps Live
Architectural Summary
This official live series explores the paradigm shift from traditional CI/CD pipelines to autonomous, agent-driven operations (Agentic DevOps). It covers technical deep-dives into Azure SRE Agents for automated root cause analysis and proactive reliability, GitHub Copilot App Mod Agents for modernizing legacy systems at scale, and AI-powered workflows across GitHub and Azure DevOps. Essential for platform engineers designing self-healing environments and scalable human-in-the-loop AI governance in 2026.