https://rudratechie.xyz/blogAI, benchmarks, and building useful things.enhttps://rudratechie.xyz/blog/code-agent-benchmarks-miss-the-hard-parthttps://rudratechie.xyz/blog/code-agent-benchmarks-miss-the-hard-partPassing coding tasks is not the same as safely shipping changes. A practical look at what agent benchmarks reveal, what they leave out, and how to test a coding agent yourself.Sat, 17 Oct 2026 06:30:00 GMTBenchmarkshttps://rudratechie.xyz/blog/rag-vs-long-context-choose-by-failurehttps://rudratechie.xyz/blog/rag-vs-long-context-choose-by-failureA clear engineering guide to retrieval-augmented generation and long context windows, with a decision framework based on freshness, corpus size, citations, and failure cost.Sat, 17 Oct 2026 06:30:00 GMTBuildinghttps://rudratechie.xyz/blog/write-technical-posts-without-ai-slophttps://rudratechie.xyz/blog/write-technical-posts-without-ai-slopA practical workflow for writing original technical content that uses AI without losing your examples, judgement, evidence, or natural voice.Sat, 17 Oct 2026 06:30:00 GMTAI & culturehttps://rudratechie.xyz/blog/measure-the-real-cost-of-ai-workhttps://rudratechie.xyz/blog/measure-the-real-cost-of-ai-workMeasure AI project cost in retries, review, latency, maintenance and errors. A practical worksheet to decide whether a feature is earning its complexity.Sat, 17 Oct 2026 06:30:00 GMTBuildinghttps://rudratechie.xyz/blog/linkedin-ai-slop-brain-rothttps://rudratechie.xyz/blog/linkedin-ai-slop-brain-rotWhy synthetic polish spreads on LinkedIn, how it changes what we reward, and a practical way to make AI-assisted writing sound like you again.Sat, 17 Oct 2026 06:30:00 GMTAI & culturehttps://rudratechie.xyz/blog/benchmark-score-is-not-a-verdicthttps://rudratechie.xyz/blog/benchmark-score-is-not-a-verdictHow to read AI benchmarks without being fooled by a leaderboard. Learn to inspect task design, scoring, baselines, contamination, variance, and the gap between a score and your job.Sat, 17 Oct 2026 06:30:00 GMTBenchmarkshttps://rudratechie.xyz/blog/benchmark-contamination-explainedhttps://rudratechie.xyz/blog/benchmark-contamination-explainedWhat training data leakage means for AI benchmark scores, what contamination can and cannot prove, and how to evaluate models on genuinely fresh work.Sat, 17 Oct 2026 06:30:00 GMTBenchmarkshttps://rudratechie.xyz/blog/how-to-start-learning-aihttps://rudratechie.xyz/blog/how-to-start-learning-aiA practical AI learning path from Python and machine learning basics to language models, evaluation and a first useful project, with a weekly study loop.Sat, 17 Oct 2026 06:30:00 GMTLearninghttps://rudratechie.xyz/blog/build-your-first-personal-ai-projecthttps://rudratechie.xyz/blog/build-your-first-personal-ai-projectA step-by-step framework to choose a personal AI project, define its success measure, build a safe prototype and learn from the results.Sat, 17 Oct 2026 06:30:00 GMTBuildinghttps://rudratechie.xyz/blog/validate-startup-problem-before-buildinghttps://rudratechie.xyz/blog/validate-startup-problem-before-buildingA practical first step for starting a startup: speak to potential users, test whether the problem happens often, measure current workarounds and avoid mistaking compliments for demand.Sat, 17 Oct 2026 06:30:00 GMTBuildinghttps://rudratechie.xyz/blog/when-to-automate-a-workflowhttps://rudratechie.xyz/blog/when-to-automate-a-workflowDecide whether to automate a repeated task using frequency, stability, error cost and human judgement. Includes a practical automation scorecard and safe rollout approach.Sat, 17 Oct 2026 06:30:00 GMTBuildinghttps://rudratechie.xyz/blog/claude-code-bmad-method-setuphttps://rudratechie.xyz/blog/claude-code-bmad-method-setupA practical setup for using Claude Code with the current BMAD-METHOD: install versioned workflows, give the agent a clear project brief, and keep agile planning proportional.Sat, 17 Oct 2026 06:30:00 GMTBuildinghttps://rudratechie.xyz/blog/context-engineering-for-coding-agentshttps://rudratechie.xyz/blog/context-engineering-for-coding-agentsA practical guide to repository instructions, task context, source-of-truth documents, code discovery and verification for coding agents like Claude Code.Sat, 17 Oct 2026 06:30:00 GMTBuildinghttps://rudratechie.xyz/blog/build-a-practical-ai-evaluation-sethttps://rudratechie.xyz/blog/build-a-practical-ai-evaluation-setCreate a trustworthy first evaluation set for an AI project: collect representative examples, write success criteria before comparing models, score failures and retain regression tests.Sat, 17 Oct 2026 06:30:00 GMTBenchmarkshttps://rudratechie.xyz/blog/tokens-context-and-transformers-for-beginnershttps://rudratechie.xyz/blog/tokens-context-and-transformers-for-beginnersA beginner-friendly explanation of tokens, context windows, transformer attention and next-token prediction, with examples that help debug real language-model applications.Sat, 17 Oct 2026 06:30:00 GMTLearninghttps://rudratechie.xyz/blog/how-to-read-an-ai-research-paperhttps://rudratechie.xyz/blog/how-to-read-an-ai-research-paperA structured reading workflow for AI research papers: identify the research question, inspect methods and baselines, read result tables carefully and find limitations before trusting the headline.Sat, 17 Oct 2026 06:30:00 GMTLearninghttps://rudratechie.xyz/blog/safe-first-deployment-for-ai-projectshttps://rudratechie.xyz/blog/safe-first-deployment-for-ai-projectsA launch checklist for a small AI feature covering user confirmation, output validation, data handling, monitoring, fallback behaviour and regression tests.Sat, 17 Oct 2026 06:30:00 GMTBuildinghttps://rudratechie.xyz/blog/deepseek-harness-cordis-and-agent-engineeringhttps://rudratechie.xyz/blog/deepseek-harness-cordis-and-agent-engineeringA technical look at DeepSeek Harness, Cordis, its plugin architecture, traceable sessions and runtime modes, plus what builders should examine before calling it the future of agents.Sat, 17 Oct 2026 06:30:00 GMTResearchhttps://rudratechie.xyz/blog/gemini-4-argon-benchmarks-and-real-workhttps://rudratechie.xyz/blog/gemini-4-argon-benchmarks-and-real-workGoogle reports strong Gemini 4 Argon results, but what do the benchmark suites measure, which gains transfer to real work, and what does a different model cost?Sat, 17 Oct 2026 06:30:00 GMTResearchhttps://rudratechie.xyz/blog/phonon-2-open-speech-recognition-on-devicehttps://rudratechie.xyz/blog/phonon-2-open-speech-recognition-on-deviceA close look at Phonon-2's compact weights, reported word error rate, Apple Neural Engine speed, licensing and the real evaluation questions for local transcription.Sat, 17 Oct 2026 06:30:00 GMTResearchhttps://rudratechie.xyz/blog/openai-ai-mathematics-results-october-2026https://rudratechie.xyz/blog/openai-ai-mathematics-results-october-2026How to assess OpenAI's October 2026 mathematics results: what the Lean formalizations establish, what the human-written manuscripts still need, and why discovery is not the same as verification.Sat, 17 Oct 2026 06:30:00 GMTResearchhttps://rudratechie.xyz/blog/speech-recognition-wer-production-evaluationhttps://rudratechie.xyz/blog/speech-recognition-wer-production-evaluationUnderstand how WER works, which transcript errors matter in practice, and how to build a speech recognition evaluation using your actual microphones, accents and vocabulary.Sat, 17 Oct 2026 06:30:00 GMTBenchmarkshttps://rudratechie.xyz/blog/reproducible-model-comparison-cost-latencyhttps://rudratechie.xyz/blog/reproducible-model-comparison-cost-latencyA repeatable method for comparing AI models on your own tasks, controlling prompts and tools while tracking success, retries, latency and total operating cost.Sat, 17 Oct 2026 06:30:00 GMTBenchmarks