{"version":"https:\/\/jsonfeed.org\/version\/1","title":"Bharadwaj Pendyala","home_page_url":"https:\/\/bharad.dev","description":"Terminal-inspired developer portfolio showcasing full-stack engineering projects, technical expertise, and software development insights. Built with Next.js, TypeScript, and modern web technologies.","author":{"name":"Bharadwaj Pendyala","url":"https:\/\/bharad.dev"},"items":[{"id":"https:\/\/bharad.dev\/blog\/ai-evals-interview","content_html":"A high eval score can hide a broken test. Hamel Husain's seven-part interview structure shows how to inspect the data, explain experiments, and make a product decision as engineering and data science roles overlap.","url":"https:\/\/bharad.dev\/blog\/ai-evals-interview","title":"Notes from Hamel Husain's experience designing eval interviews","date_modified":"2026-10-12T00:00:00.000Z","tags":["AI","Evals","Engineering","Career"]},{"id":"https:\/\/bharad.dev\/blog\/build-the-eval-pyramid","content_html":"Reproduce the refund agent that said \"refunded\" when it hadn't, watch a ledger check and a critical gate block it, then fix the bug and rerun. A walkthrough of the open-source Eval Pyramid Starter, one failure at a time.","url":"https:\/\/bharad.dev\/blog\/build-the-eval-pyramid","title":"The Eval Pyramid, Part III: Trace One Failure from Fixture to Release Gate","date_modified":"2026-08-10T00:00:00.000Z","tags":["AI","Engineering","Agents","Evals","Testing","TypeScript"]},{"id":"https:\/\/bharad.dev\/blog\/measuring-agent-reliability","content_html":"One passing trial says a system can succeed. Repeated trials, computed per task, say whether people can depend on it. Here's how to read pass@1, pass@k and pass^k, why the same 75% can describe two opposite systems, and how to check a model judge before it gates a release.","url":"https:\/\/bharad.dev\/blog\/measuring-agent-reliability","title":"The Eval Pyramid, Part II: Measuring Agent Reliability with pass@k, pass^k, and Judges You Can Trust","date_modified":"2026-08-09T00:00:00.000Z","tags":["AI","Engineering","Agents","Evals","Testing"]},{"id":"https:\/\/bharad.dev\/blog\/eval-pyramid","content_html":"An eval pyramid orders graders by what one verdict costs: code checks at the base, calibrated model judges above them, people at the top. Repeated trials are a separate axis that applies to every layer, and keeping them apart makes both easier to reason about.","url":"https:\/\/bharad.dev\/blog\/eval-pyramid","title":"The Eval Pyramid, Part I: Testing Systems That Don't Behave the Same Way Twice","date_modified":"2026-08-08T00:00:00.000Z","tags":["AI","Engineering","Agents","Evals","Testing"]},{"id":"https:\/\/bharad.dev\/blog\/rag-retrieval-augmented-generation","content_html":"A chat model only knows what it saw in training. RAG is the plumbing that lets it answer from your private documents, today's news, or anything it was never trained on: a second kind of model that turns text into coordinates, a database of those coordinates, and a search step that runs before the LLM ever sees your prompt.","url":"https:\/\/bharad.dev\/blog\/rag-retrieval-augmented-generation","title":"RAG: How an LLM Answers From Documents It Never Saw","date_modified":"2026-07-03T00:00:00.000Z","tags":["AI","ML","Learning","RAG","LLMs"]},{"id":"https:\/\/bharad.dev\/blog\/from-transformer-to-llm","content_html":"A stack of transformer layers is not yet ChatGPT or Claude. This is the rest of the path: how text becomes tokens, how a raw next-word predictor turns into an assistant across three training phases, how LoRA customizes a model on a budget, and why everyone is racing for data and compute.","url":"https:\/\/bharad.dev\/blog\/from-transformer-to-llm","title":"From Transformer to ChatGPT: The Part That Isn't the Architecture","date_modified":"2026-06-25T00:00:00.000Z","tags":["AI","ML","Learning","Transformers","LLMs"]},{"id":"https:\/\/bharad.dev\/blog\/transformers-and-attention","content_html":"We use models built on transformers every day, yet the architecture itself usually stays a black box. It doesn't have to, and following it doesn't take heavy math. A visual, ground-up walkthrough: how words become numbers, how attention lets them shape each other, and how it all turns into a next-word prediction.","url":"https:\/\/bharad.dev\/blog\/transformers-and-attention","title":"Transformers and Attention: A Plain-English Guide for Engineers","date_modified":"2026-06-24T00:00:00.000Z","tags":["AI","ML","Learning","Transformers"]},{"id":"https:\/\/bharad.dev\/blog\/cognitive-surrender","content_html":"Reading notes from Addy Osmani on the difference between using AI as a tool and letting it quietly replace your thinking. Key research findings and heuristics worth keeping.","url":"https:\/\/bharad.dev\/blog\/cognitive-surrender","title":"Notes on Cognitive Surrender","date_modified":"2026-05-16T00:00:00.000Z","tags":["AI","Engineering","Learning"]},{"id":"https:\/\/bharad.dev\/blog\/why-i-joined-deep-atlas-accelerator","content_html":"I've been shipping with AI tools for a year. At some point the gap between using them and actually understanding them stopped feeling fine. This is about what I decided to do about it.","url":"https:\/\/bharad.dev\/blog\/why-i-joined-deep-atlas-accelerator","title":"Why I Joined the Deep Atlas AI\/ML Accelerator","date_modified":"2026-05-09T00:00:00.000Z","tags":["AI","ML","Learning"]},{"id":"https:\/\/bharad.dev\/blog\/navigating-ai-careers","content_html":"The AI job market is noisy and competitive. Here are the ideas worth keeping from a talk by Andrew Ng and Laurence Moroney on careers, technical debt, and what it takes to ship real AI work.","url":"https:\/\/bharad.dev\/blog\/navigating-ai-careers","title":"Navigating the AI Job Market: What Actually Matters","date_modified":"2026-04-04T00:00:00.000Z","tags":["AI","Career","Learning"]}]}