How do you set durable DevAI goals given the complexity and rapid change in the space? The good news is, distinct “styles” of DevAI are becoming evident, providing a basis for setting expectations and goals. More on this idea:

Kind of a wild observation from @andy_pavlo speaking on a @DBOS_Inc webinar: a year ago LLMs couldn’t do most of the assignments in his database course, but now they can solve most. An optimistic data point for AI taking on harder coding tasks.

AI coding agents are moving from hype to hands-on. ⚡️ Join @rbren_dev (@allhands_ai ) & @AJFerrari (@_jellyfish_co ) for a candid look at how agentic AI is transforming real developer workflows — from refactoring to team dynamics and leadership strategy. #AI #Engineering

For more than 10 years, the DORA report has shaped engineering leadership trends and metrics for an entire industry. Now, as a Gold sponsor, Jellyfish is bringing the report’s insights to you first in an exclusive webinar on October 30, featuring the report’s lead author, Nathen

Last month at the ELC annual conference, Krishna Kannan shared with attendees the AI Impact Framework - built in collaboration with 300+ engineering leaders and grounded in data from over 20,000 developers. Now, you can watch the recording here: https://t.co/k9kjWrfoOP And

The new DORA report is out: https://t.co/9PK45X11P9! Come join DORA lead Nathen Harvey, myself, and the @_jellyfish_co team on Oct 30 to hear about the new insights that DORA has uncovered, and how to put these into action:

@MLB You can look further back for even more balanced years, but in the past quarter century, I think those are the ones.

With a standard deviation of winning percentages around 0.144, that makes this an even more balanced year than 2014 (at 0.146 SD), but a little less balanced than 2007 (0.138 SD).

Closing in on the end of the @MLB season, there’s only one team over .600 (Brewers, barely) and two teams under .400 (Rockies and White Sox). A “balance of powers” kind of year.

Is the “Great Flattening” in tech going to delete engineering management?

More and more Engineering leaders I speak with have quickly graduated from “How do we get started with AI coding tools?” to the tougher question of “How do we actually measure the impact?” My latest post looks at this question

Software is going through a massive transformation with AI coding. What will it all mean for professional software engineering? Jellyfish has been doing some amazingly interesting research on this. Come learn about it at GLOWLive on April 17

I voted in the 2025 Cornell Alumni Trustee Election! You can, too! Choose your candidates and cast your vote between February 1-28. Your vote matters! 🐻 https://t.co/LiuvMHgYZ3 via @CornellAlumni

Just a little double unicorn day here @Salsify

Whenever I’m cutting up and boiling a large Rutabaga it definitely feels a bit like a Star Wars food. Like Aunt Beru should probably be doing this. Wash down with some nice blue milk.

One of our amazing @salsify engineers, Erik, presenting at #GraphQLConf2021 right now, highlighting two of the highest sentiment elements of our dev stack: #GraphQL and TypeScript. Really highlights the continuing importance and value of typed languages.

This should be a great event for anyone working with #GraphQL. We've got a nice overview of our approach to tooling at @salsify, and a bunch of other great orgs are presenting. https://t.co/dxStP5rhUS #GraphQLConf2021

However the organization gets there, Engineering is accountable for clearly identifying situational taxes on efficiency (such as technical debt) and working with the business to chart different courses forward.

Sometimes debt challenges may be the result of poor decisions in the Eng org, but more typically it’s a cross-functional issue created by the whole business - the age-old story of “we had to go fast and iterate to get product-market fit, and couldn’t find time to manage debt.”

It’s tempting to blame your Engineering team for this, but it really isn’t a function of the capability of the engineers involved, nor of the quality of the Eng org. It’s a function of history and the resulting technical environment, and can be expensive and slow to “fix.”.