# Arjun Prabhulal > Personal website of Arjun Prabhulal, a Google Developer Expert writing about AI agents, retrieval, and cloud engineering. The homepage features the aiml.sh course catalog. Articles are available at /blog/; course cards link to aiml.sh, and the whatisai link below the masterclass opens its explainers and news. The original Google ADK 1.x masterclass remains on this website. ## Main pages - [Blog](https://arjunprabhulal.com/blog/): Technical articles and tutorials. - [About](https://arjunprabhulal.com/about/): Background, speaking, and projects. - [Courses](https://arjunprabhulal.com/): aiml.sh course catalog, a separate whatisai learning path, and the original ADK masterclass. - [Projects](https://arjunprabhulal.com/projects/): Personal projects. - [Speaking](https://arjunprabhulal.com/sessions/): Conference and meetup talks. - [Contact](https://arjunprabhulal.com/contact/): Contact page. ## aiml.sh courses Catalog snapshot: 17 courses, 60 named modules, 224 lessons, and approximately 89 reading hours. Lab time is additional. Snapshot date: 2026-10-27. The live catalog is the authority for subsequent changes. - [Browse aiml.sh courses](https://aiml.sh/courses/) - [AI Fundamentals](https://aiml.sh/courses/ai-fundamentals/): 11 lessons; about 2 hours of reading. - [Model Context Protocol (MCP)](https://aiml.sh/courses/mcp/): 7 lessons; about 2 hours of reading. - [Building Agents with Google ADK 2.0](https://aiml.sh/courses/google-adk/building-agents/): 17 lessons; about 6 hours of reading. - [Connecting Agents to Tools and Services with Google ADK 2.0](https://aiml.sh/courses/google-adk-integrations/): 6 lessons; about 2 hours of reading. - [Running and Deploying Agents with Google ADK 2.0](https://aiml.sh/courses/google-adk-production/): 20 lessons; about 8 hours of reading. - [Building Live and Voice Agents with Google ADK 2.0](https://aiml.sh/courses/google-adk-live/): 7 lessons; about 3 hours of reading. - [Building Agents with LangGraph](https://aiml.sh/courses/langgraph/): 9 lessons; about 4 hours of reading. - [Retrieval-Augmented Generation (RAG)](https://aiml.sh/courses/rag/): 6 lessons; about 2 hours of reading. - [Building Agentic RAG Applications](https://aiml.sh/courses/agentic-rag/): 37 lessons; about 15 hours of reading. - [LLM Evals](https://aiml.sh/courses/llm-evaluation/): 6 lessons; about 2 hours of reading. - [Building, Evaluating, and Deploying Agents with Google Agents CLI](https://aiml.sh/courses/agents-cli/): 6 lessons; about 2 hours of reading. - [Building AI Systems](https://aiml.sh/courses/building-ai-systems/): 21 lessons; about 9 hours of reading. - [Building MLOps](https://aiml.sh/courses/building-mlops/): 21 lessons; about 10 hours of reading. - [Build a Small Language Model From Scratch](https://aiml.sh/courses/llm-from-scratch/): 11 lessons; about 5 hours of reading. - [Agent Design, Optimization and Harnesses](https://aiml.sh/courses/agent-engineering/): 9 lessons; about 4 hours of reading. - [Context and Memory Engineering for AI Agents](https://aiml.sh/courses/context-and-memory/): 10 lessons; about 5 hours of reading. - [LLM Inference Engineering](https://aiml.sh/courses/llm-inference/): 20 lessons; about 8 hours of reading. ## Other learning paths - [Learn AI, five minutes at a time](https://whatisai.fyi/start/): Understand AI in plain words, with clear diagrams and sources. Follow 29 short chapters across fundamentals, deep learning, language models, apps and agents, and production, evaluation, and safety. - [Google ADK Masterclass (ADK 1.x)](https://arjunprabhulal.com/courses/adk-masterclass-hands-on-series/): Original 20-module series. ## Articles hosted on this website - [How I Built Semantic Search Over 300 Million Plus Academic Papers](https://arjunprabhulal.com/how-i-built-semantic-search-300m-academic-papers/): What I learned building semantic paper search for Rivul AI on one machine: a retrieval funnel of BGE-small, FAISS IVF-PQ, INT8 re-scoring and title matching, plus ranking, filters and read-time fixes. - [Four Layers of Agent Engineering](https://arjunprabhulal.com/graph-harness-loop-context/): Most agent bugs get blamed on the model. Most of them actually live in one of four layers: context, loop, harness, or graph. Here's how I tell them apart. - [Top 30 AI/ML Terms Every Developer Should Know](https://arjunprabhulal.com/ai-ml-essential-terms/): A comprehensive glossary of 30 must-know AI and Machine Learning terms. From neural networks to RAG, transformers to agentic AI - explained simply. - [Google ADK - MCP Toolbox for Databases (PostgreSQL)](https://arjunprabhulal.com/adk-mcp-toolbox/): Connect ADK agents to PostgreSQL databases using MCP Toolbox for enterprise-grade data access. - [Google ADK - Events](https://arjunprabhulal.com/adk-events/): Understand the event system for real-time streaming and debugging. - [Google ADK - Artifacts](https://arjunprabhulal.com/adk-artifacts/): Store and manage files, images, and data produced by agents. - [Google ADK - Callbacks](https://arjunprabhulal.com/adk-callbacks/): Intercept and customize agent behavior at various execution points. - [Google ADK - Context Management](https://arjunprabhulal.com/adk-context-management/): Optimize agent performance with context caching and compression techniques. - [Google ADK - Model Context Protocol (MCP)](https://arjunprabhulal.com/adk-mcp-deep-dive/): Understanding the Model Context Protocol - architecture, connection types, and production patterns. - [Google ADK - Session, State & Memory](https://arjunprabhulal.com/adk-sessions-state/): Manage conversation history, state, and memory for multi-turn interactions. - [Google ADK - Third Party Tools (GitHub, Firecrawl)](https://arjunprabhulal.com/adk-third-party-tools-github/): Connect your agents to external services for version control and web content extraction. - [Google ADK - Multi-Tool Agents](https://arjunprabhulal.com/adk-multi-tool-agent/): Build agents that leverage multiple tools together to accomplish complex tasks. - [Google ADK - OpenAPI Tools with Authentication](https://arjunprabhulal.com/adk-custom-tools-openapi/): Transform OpenAPI-defined APIs into agent tools and configure authentication for secure access. - [Google ADK - Custom Function Tools](https://arjunprabhulal.com/adk-custom-tools-function/): Build custom functions that add specialized capabilities to your agents beyond built-in tools. - [Google ADK - Built-in Tools (Vertex AI Search)](https://arjunprabhulal.com/adk-builtin-tools-vertex-search/): Build agents that can search across your enterprise data using the Vertex AI Search built-in tool. - [Google ADK - Built-in Tools (Vertex AI RAG Engine)](https://arjunprabhulal.com/adk-builtin-tools-rag/): Learn how to use the Vertex AI RAG Engine built-in tool to create grounded agents with your own data. - [Google ADK - Built-in Tools (Google Search, Code Executor)](https://arjunprabhulal.com/adk-built-in-tools/): Explore the Google Search and Code Executor built-in tools. - [Google ADK - Multi-Agent Systems](https://arjunprabhulal.com/adk-multi-agent-systems/): Compose multiple agents into hierarchical systems for complex applications using agent composition patterns. - [Google ADK - Workflow Agents (Sequential, Loop, Parallel)](https://arjunprabhulal.com/adk-workflow-sequential-loop-parallel/): Create dynamic task flows with sequential, loop, and parallel workflows in Google ADK. - [Google ADK - Building LLM Agents](https://arjunprabhulal.com/adk-llm-agents/): Build intelligent, model-powered agents with Google ADK. - [Google ADK - Setting Up Agents (CLI, Web, Programmatic)](https://arjunprabhulal.com/adk-setting-up-agent/): Learn how to set up and run ADK agents using three different methods: command-line interface, web interface, and programmatic Python approach. - [Google ADK - Getting Started](https://arjunprabhulal.com/adk-getting-started/): Introduction to ADK – Overview, environment setup, and your first agent. - [Google ADK - Visual Agent Builder](https://arjunprabhulal.com/google-adk-visual-agent-builder/): Google ADK Visual Agent Builder to build and manage AI agents without coding - [RAG Simplified with Google Gemini API's File Search Tool](https://arjunprabhulal.com/rag-simplified-with-filesearch-tool/): RAG implementation using Gemini API's File Search tool ## Projects - [aiml.sh](https://arjunprabhulal.com/projects/aiml/): Hands-on AI engineering courses, runnable Python examples, and evaluations. [Visit aiml.sh](https://aiml.sh/). - [Rivul AI](https://arjunprabhulal.com/projects/rivul-ai/): AI research workspace. [Visit Rivul](https://rivul.ai/). - [superPDF AI](https://arjunprabhulal.com/projects/superpdf/): PDF tools and document workflows. [Visit superPDF](https://superpdf.ai/). - [whatisai](https://arjunprabhulal.com/projects/whatisai/): AI explainers and daily news. [Visit whatisai](https://whatisai.fyi/). ## AI Daily News - [Today](https://whatisai.fyi/news/) - [Models](https://whatisai.fyi/news/models/) - [Open source](https://whatisai.fyi/news/open-source/) - [Research](https://whatisai.fyi/news/research/) - [Trending](https://whatisai.fyi/news/trending/) - [GitHub trending](https://whatisai.fyi/news/github/) ## Feeds and contact - [RSS feed](https://arjunprabhulal.com/feed.xml) - [Sitemap](https://arjunprabhulal.com/sitemap.xml) - [Security contact](https://arjunprabhulal.com/.well-known/security.txt) - [LinkedIn](https://www.linkedin.com/in/arjun-prabhulal/) - [GitHub](https://github.com/arjunprabhulal) - [Medium](https://medium.com/@arjun-prabhulal) - [X](https://x.com/ArjunPrabhulal)