Language models · 13 min read
How Agentic RAG Learns to Say "I Don't Know"
A model that doesn't know your company's policies will make them up. RAG hands it the right paragraph, and agentic RAG makes it check that paragraph before it answers.
0xKishan, the latest posts by Kishan Kumar
Also in this issue
- Language models · 10 min
What Are Word Embeddings, Anyway?
Meaning is not typed in by anyone. It falls out of a prediction game played over billions of sentences, and the leftover geometry is what we call an embedding.
- Language models · 12 min
What Are Tokens in an LLM, Anyway?
The algorithm behind GPT's tokenizer is called BPE, byte pair encoding.
- Language models · 6 min
Don't Let AI Make You Artificial
We use LLMs in almost every part of our life.
- Distributed systems · 25 min
Designing The Webhook Delivery System
A webhook is a lightweight, event-driven mechanism that allows one application to automatically send real-time data to another via HTTP (typically a POST request) immediately when a specific event occurs.
By topic
All 65 posts →- Oct 17, 2026 · 13 min
How Agentic RAG Learns to Say "I Don't Know"
- Jul 31, 2026 · 10 min
What Are Word Embeddings, Anyway?
- Jul 25, 2026 · 12 min
What Are Tokens in an LLM, Anyway?
- Feb 23, 2026 · 25 min
Designing The Webhook Delivery System
- Feb 22, 2026 · 30 min
Design a News Aggregator System
- Oct 15, 2025 · 8 min
Before You Add Another Service for Caching, Read This.
- Mar 16, 2025 · 19 min
Build your own API Gateway from Scratch in Java
- Jul 27, 2024 · 20 min
Designing and Implementing a URL Shortener in Java
- Jul 11, 2024 · 17 min
Designing and Implementing an API Rate Limiter in Java
- Dec 21, 2023 · 6 min
Chainlink: Bridging the On-Chain and Off-Chain Worlds
- Nov 18, 2023 · 7 min
Introduction to Vert.x
- Sep 23, 2023 · 6 min
MEV Explained: How Miners and Validators Extract Extra Value from Blockchain Transactions
About the writer
Kishan Kumar
Senior Software Engineer · PhonePe
Senior software engineer at PhonePe. I write long explainers on distributed systems, language models, Java and blockchains.
I've spent over six years building systems at scale. At PhonePe I lead the logging charter, which handles around 65 TB of logs a day.
Read more →Currently reading
- 2017 · Vaswani et al.
Attention Is All You Need
- 2003 · Ghemawat, Gobioff, Leung
The Google File System
- 2007 · DeCandia et al.
Dynamo: Amazon's Highly Available Key-value Store
- 2011 · Yuval Noah Harari
Sapiens: A Brief History of Humankind