@vaishnavi-sharma

Kubernetes

Dec 2, 2025 · 5 min read · As developers, we write a lot of code — Java, PHP, Node.js, anything. But developing the application is only one challenge; deploying it in a reliable and consistent manner is another. Whatever works on our local machine should also work the same way...

Cosine Similarity, Normalization, FAISS, LangChain, LangGraph and CLIP Internals

Nov 26, 2025 · 3 min read · This blog covers the core technical foundations you need when implementing RAG systems with images, text, vector stores, and agent workflows. Cosine Similarity Cosine similarity measures how similar two vectors are based on their direction, not thei...

Understanding RAG (Retrieve, Augment, Generate) With Multimodal Embeddings

Nov 26, 2025 · 2 min read · RAG has three important steps: Retrieve, Augment, and Generate.This is the fundamental pipeline used in most LLM applications today. 1. Retrieve Retrieve means finding the most relevant data from your knowledge source. Sources can be: Text document...

Design a System That Lets Users Upload and View Profile Pictures

Nov 25, 2025 · 3 min read · This system must allow users to upload their profile picture, validate the file, store it safely, process it into multiple sizes, and make it load quickly anywhere in the world.Here is the complete design. Functional Requirements Users should be ab...

System Design Basics — Part 2

Nov 20, 2025 · 4 min read · Scaling in real systems is not done randomly. There are clear rules and policies.Scaling is based on parameters such as: If requests go beyond 1000, add a server If CPU usage reaches 70%, spin up another server Every company uses scaling differen...