Medium

Rahul Gite

Pinned

In

Level Up Coding

by

Rahul Gite

·

1d ago

An Avocado, a Microphone, and a Different Way to Work With Codex

My experiments with making coding agents easier to understand, collaborate with, and trust during everyday development.

An Avocado, a Microphone, and a Different Way to Work With Codex

Pinned

In

Level Up Coding

by

Rahul Gite

·

Oct 21

The Code Was Fast. Fixing It Wasn’t: My Experience With Spec-Driven Development

My journey from repetitive prompting to spec-driven development, and what I learned about getting AI to write better production code.

The Code Was Fast. Fixing It Wasn’t: My Experience With Spec-Driven Development

Pinned

In

Level Up Coding

by

Rahul Gite

·

Sep 28

What Google’s New SDLC Paper Got Right About AI-Assisted Software Development

As coding agents make implementation cheaper, the real work is slowly moving towards specification, context, verification and judgment.

What Google’s New SDLC Paper Got Right About AI-Assisted Software Development

Pinned

In

Level Up Coding

by

Rahul Gite

·

Jul 21, 2025

Inside RAG Systems: The Anatomy of Retrieval-Augmented Generation

What makes RAG systems tick? A high-level breakdown of components, workflows, and why RAG is essential for grounded language generation.

Inside RAG Systems: The Anatomy of Retrieval-Augmented Generation

In

Level Up Coding

by

Rahul Gite

·

Nov 2

Six Agent Harness Techniques That Actually Worked for Me

What building production AI agents taught me about context engineering, routing, cost optimisation, human intervention, and evaluation.

Six Agent Harness Techniques That Actually Worked for Me

In

Level Up Coding

by

Rahul Gite

·

Oct 21

Google’s TabFM: Can We Finally Do Machine Learning Without Training a Model?

Google’s new tabular foundation model brings zero-shot predictions to structured data.

Google’s TabFM: Can We Finally Do Machine Learning Without Training a Model?

In

Level Up Coding

by

Rahul Gite

·

Oct 12

Not Every AI Problem Needs a Description: Understanding Jev and System One Models

How a new class of AI models is focusing on fast decisions, calibrated confidence, and the smaller tasks inside our AI applications.

Not Every AI Problem Needs a Description: Understanding Jev and System One Models

In

Level Up Coding

by

Rahul Gite

·

Oct 7

The Last Mile of RAG: From Retrieved Chunks to a Trustworthy Answer

Retrieval gives us evidence. But building a good RAG system also means deciding whether that evidence is enough, generating only from it…

The Last Mile of RAG: From Retrieved Chunks to a Trustworthy Answer

In

Level Up Coding

by

Rahul Gite

·

Sep 21

Reranking for better relevance

Retrieving relevant documents is only half the job. Reranking helps us identify which results actually deserve to become context.

Reranking for better relevance

In

Level Up Coding

by

Rahul Gite

·

Dec 30, 2025

Retrieving the Right Stuff: Context Extraction in RAG

Precision matters. Strategies for embedding-based retrieval, chunking, and relevance scoring to extract useful context.

Retrieving the Right Stuff: Context Extraction in RAG

Rahul Gite

Rahul Gite

I love to write about anything new that I learn from my work and in general. Lets connect on LinkedIn: https://www.linkedin.com/in/rahul-gite-connect/

Following

Help

Status

About

Careers

Press

Blog

Store

Privacy

Rules

Terms

Text to speech