The Pragmatic Engineer

The Pragmatic Engineer is Gergely Orosz’s engineering newsletter that brings an engineer-first, data-grounded lens to how AI is changing software teams, tooling, and hiring.

Where the hands-on voices show you how to use an agent, The Pragmatic Engineer shows you how a whole engineering organization absorbs one, and that org view is what makes it distinct in this category.

What it is

#

A long-running engineering newsletter (and podcast) by Gergely Orosz, author of The Software Engineer’s Guidebook. Its AI coverage is systematic: tooling adoption surveys, printed case studies of teams using coding agents, and interviews with practitioners like Steve Yegge on agent workflows. It consistently runs the data-gathering (AI tooling surveys) that individual tool blogs cannot, giving it a rare empirical angle on adoption.

Status

#

Active and heavily subscribed, with more than 1,100,000 subscribers as of 2026-09-22. As of 2026-10-02 it runs the AI tooling-for-2026 surveys and current agent-adoption coverage: The Pulse on the Ruby-on-Rails creator sparking a new “death of coding by hand” debate (2026-09-24), How will AI change operating systems? Part 2: Windows, on making the OS AI-agent-friendly (2026-09-22), AI Skills with Matt Pocock (2026-09-17), Inside OpenAI’s agentic software factory (2026-09-15), Building Codex with Tibo Sottiaux (2026-09-09), and What is happening with code reviews? (2026-09-08). A large share of AI-specific content sits behind the paid tier, which funds the survey and interview work.

Strengths

#

• The org lens is unique in this category: hiring, tooling budgets, and team structure around agents.
• Data-backed pieces, especially the annual AI tooling surveys, quantify adoption rather than anecdote.
• Interview range is broad and on-topic: Yegge on agent futures, Kent Beck on engineering, context-engineering deep dives.
• High editorial polish and packed post-mortems that survive as references.

Cautions

#

• The engineering-leadership focus is weaker on the day-to-day terminal-and-token details of running agents.
• Much of the best AI material is paywalled, so the free tier underrepresents the substance.
• Orosz profits from the newsletter business model itself, which occasionally surfaces in sponsorship and event promotion.

Pricing

#

Freemium Substack. A substantial and growing share of AI-engineering content is paid; the free tier retains enough volume to evaluate the fit.

Compared to

#

• Simon Willison: the individual practitioner’s daily notes versus the org-and-industry analysis; read both and you get the loop and the company simultaneously.
• Latent Space: Latent Space for the AI-native startup and research view, Pragmatic Engineer for the established-engineering-company view.
• Andrej Karpathy: Karpathy sets the technical vocabulary, Orosz documents how enterprises actually land it.

Bottom line

#

Recommended for engineering leaders and senior engineers who need the data- and org-grounded view of how teams adopt agents. Not for someone chasing pure hands-on tool detail who resents a paywall.

Top 5 recommended reading

#

• AI Tooling for Software Engineers in 2026 - The annual survey that quantifies what engineers actually use, the data backbone of the newsletter’s AI coverage.
• How Claude Code is built - His interview with Boris Cherny, the primary account of how the dominant coding agent was designed and scales as a product.
• Steve Yegge on AI Agents and the Future of Software Engineering - The maximalist agent-future argument in long form, the position his adoption data tests against.
• Inside OpenAI’s agentic software factory - His September 2026 deep dive into an AI-first engineering organization, the org view that defines the newsletter.
• The AI Engineering Stack - His framing with Chip Huyen of how the AI engineering layers fit together, the map for the rest of the coverage.

Changes

#

• 2026-08-29 - Created in the People and publications category seed as the org-and-data voice, noting the paywalled core and annual AI tooling surveys.
• 2026-09-13 - Added the 1.1M subscriber figure and the September agent-adoption coverage (Codex interview, code-review report, Ramp’s Inspect).
• 2026-09-16 - Added the “Inside OpenAI’s agentic software factory” deep dive (2026-09-15) to the September agent-adoption coverage.
• 2026-09-18 - Added the “AI Skills with Matt Pocock” episode (2026-09-17) to the September agent-adoption coverage; subscriber count held at over 1,100,000.
• 2026-09-22 - Added “How will AI change operating systems? Part 2: Windows” (2026-09-22) to the September agent-adoption coverage; subscriber count held at over 1,100,000.
• 2026-09-24 - Added the Top 5 recommended reading section.

See also

#

• Model Selection for Coding Tasks - the tooling-adoption data meets the model-decision process
• Context Management Patterns - a pragmatic-engineering deep-dive topic in pattern form
• Rethinking Code Review in the Age of LLMs - the org-level review shift his case studies document
• Scaling the LLM Agent Company - org design at agent scale, the same subject as his surveys

References

#

• https://newsletter.pragmaticengineer.com/ - the newsletter home and AI tag archive
• https://newsletter.pragmaticengineer.com/feed - the RSS feed grounding the latest issues and dates
• https://newsletter.pragmaticengineer.com/p/ai-tooling-2026 - the annual AI tooling survey for software engineers
• https://newsletter.pragmaticengineer.com/p/steve-yegge-on-ai-agents-and-the - his interview with Steve Yegge on AI agents
• https://newsletter.pragmaticengineer.com/p/the-ai-engineering-stack - his framing of the AI engineering stack
• https://newsletter.pragmaticengineer.com/p/from-ides-to-ai-agents-with-steve - the IDEs-to-agents episode with Yegge
• https://www.pragmaticsummit.com/ - the summit brand behind the paid content line