ARAman Rajinblogs.amanraj.me·2d ago · 20 min readWhere RAG Fails: Understanding the Limitations of Retrieval-Augmented GenerationLarge language models can answer questions, write content, summarize information, and help with many other tasks. However, they have one important limitation: they do not automatically know everything00
ARAman Rajinblogs.amanraj.me·5d ago · 11 min readSecure Image Uploads from Expo to Amazon S3 Using Presigned URLsUploading images directly from an Expo app to Amazon S3 can make your uploads faster and reduce work on your backend server. However, you should never put AWS access keys inside your Expo app. Anyone 00
ARAman Rajinblogs.amanraj.me·Jun 18 · 12 min readHow to Run Open-Source AI Models Locally for Free on Mac and WindowsToday, you do not need to pay for an AI API just to test AI models. You can run many open-source or open-weight AI models directly on your own laptop or PC. This means the model runs on your machine, 00
ARAman Rajinblogs.amanraj.me·Jun 17 · 7 min readDeploying Qwen3 235B on JarvisLabs with vLLM This guide explains how to deploy Qwen3 235B on JarvisLabs and expose it as an OpenAI-compatible API endpoint. We will deploy this model: Qwen/Qwen3-235B-A22B-Instruct-2507-FP8 We will serve it with 00
ARAman Rajinblogs.amanraj.me·Apr 5 · 7 min readStop Committing Your .env Files: The Global Gitignore Hack Every Developer ForgetsOkay, real talk. How many times have you cloned a fresh repo, opened it in VS Code, started hacking away, only to realize—crap—you just committed your .env file with your database password in plaintex00