LinearB Blog | Insights and Best Practices for Software Engineering Leaders

Ralph loops make agentic coding reliable with ruthless context resets

HumanLayer founder Dex Horthy shares the origin story of the Ralph loop and breaks down his RPI (Research, Plan, Implement) methodology—a framework designed to...

Slack is transforming work by integrating AI agents into everyday conversations

Slack’s Kurtis Kemple explains how the platform is evolving into an agentic work operating system where human intent meets autonomous ability. He breaks down...

AI-driven code generation is not doubling productivity without SDLC transformation

LinearB CEO Ori Keren reveals why the surge in AI code generation has failed to double productivity, explaining that SDLC bottlenecks—like code reviews and...

Join our community of data-driven dev leaders

LinearB may send you email occasionally about how you can optimize productivity. We will not share your information with anyone. Ever.

Why OpenAI Codex builds the agent first, not the product

OpenAI’s Thibault Sottiaux reveals how the Codex team ruthlessly simplifies agent design by removing manual scaffolding, solving context compaction at the...

DevEx

How Spotify turned internal chaos into an industry standard

Spotify’s Tyson Singer explains how the company evolved from manual spreadsheets to Backstage, the global IDP standard, using Golden Paths to scale autonomy...

Mastering Ralph loops transforms software engineering with LLM automation

Geoffrey Huntley introduces the Ralph Wiggum Loop, a deterministic LLM pattern that manages context rot and compaction by treating the context window like...

14,000 Engineers, One Pipeline: How Capital One Standardizes for Speed

Capital One’s Ameesh Paleja explains how standardization acts as a force multiplier for 14,000 technologists, enabling AI-powered automation and a unified...

Why AI elevates engineering leadership instead of replacing it

Engineering executive Manoj Mohan explains why AI elevates leadership by shifting the focus from code generation to high-level strategy, utilizing his 3GF...

The end of pre-compute: why the future of AI is on-demand

Chalk co-founder Elliot Marx argues that pre-compute fails in real-time AI by causing latency and errors, advocating for on-demand, real-time data pipelines to...

The AI Productivity Platform

Features

Blog