Sandboxes and Policies for AI Agents: A Practical Architecture for Your First DeploymentAn agent reads a repository, edits a function, and runs the tests. Then it encounters a README instruction asking it to upload its environment variables for “debugging.” A prompt telling it not to rev11 min read·rpsarathy.hashnode.dev12MKOct 22
The RAG vs. OKF Debate Is a Category Error — and It's Hiding the Architecture You Actually NeedEver since Google Cloud published the Open Knowledge Format in June 2026, my feed has been full of "RAG vs. OKF" takes: which one wins, which one your agents should use, whether retrieval is dead. Aft9 min read·rpsarathy.hashnode.devOct 14
Jev Isn’t an LLM Replacement. It’s a Decision Plane for Agentic SoftwareJev Isn’t an LLM Replacement. It’s a Decision Plane for Agentic Software12 min read·rpsarathy.hashnode.devOct 1
Why Your RAG Pipeline Needs a Reranker: Bi-Encoders vs. Cross-Encoders, and When It MattersWhy Your RAG Pipeline Needs a Reranker: Bi-Encoders vs. Cross-Encoders, and When It Matters7 min read·rpsarathy.hashnode.dev11MSep 22
Graph Engineering: The Idea That Fixed AI AgentsEvery few months, the AI world gives the same problem a new name. Prompt engineering. Context engineering. Loop engineering. This year it's graph engineering. I get the eye-roll. The renaming gets old8 min read·rpsarathy.hashnode.devSep 17
