What Is an LLM?
A large language model predicts the next piece of text. Understanding that one mechanism explains both what it's good at and how it fails.
Beginner7 min readDaksh Bathla04
AI
Language models, coding agents, and what AI-assisted development gets right and wrong.
9 LESSONS
What Is AI-Assisted Development?
Building software with a model doing part of the typing. The skill that matters is no longer writing code — it's specifying and reviewing it.
Beginner7 min readDaksh BathlaAI-Generated Code vs Production Code
Code that runs and code you can operate for two years are different products. Here's the gap, itemised.
Intermediate7 min readDaksh BathlaHow Coding Agents Work
An agent is a language model in a loop with tools. Knowing the loop tells you exactly where it will go wrong.
Intermediate7 min readDaksh BathlaWhat AI Coding Tools Get Wrong
The specific, recurring failures — and why each one comes from how these tools work rather than from a bug someone will fix.
Intermediate7 min readDaksh BathlaPrompting for Development
Writing the request that produces code you keep. Six parts, in order, with the failure each one prevents.
Beginner8 min readDaksh BathlaContext Engineering Basics
Deciding what a model sees, and what it doesn't. On real work this matters more than prompt wording.
Intermediate8 min readDaksh BathlaDebugging AI-Generated Code
Debugging code nobody wrote by hand needs a different order of attack, because you can't rely on remembering why any line is there.
Intermediate8 min readDaksh BathlaAI API Basics
Calling a language model from your own product: the request shape, the parameters that matter, cost, and the failure modes that only appear in production.
Intermediate8 min readDaksh Bathla