Clojure in product. Would you do it again? Episode 11. Jeremiah Via, The New York Times

Episode 11 of “Clojure in Product. Would You Do It Again?” features Jeremiah Via, Staff Software Engineer at The New York Times, in conversation with Artem Barmin and Vadym Kostiuk. Jeremiah shares how Clojure was introduced across the search stack at a major US media company, highlighting JVM interop and tooling as key enablers for data-heavy workloads.

We cover:
• His Clojure origin story and migration from PHP, Erlang, Python, and Java
• How the team pushes vector embeddings into Elasticsearch for AI features
• Hiring for domain expertise over Clojure fluency, and onboarding via REPL workflows
• Production practices: memory sizing, performance tuning, and code discipline

Jeremiah also reflects on AI tooling, "Now with AI stuff, people can be productive very fast without understanding it, using a cursor and tools like that." He cautions that it remains to be seen how this will affect the deeper mental model of learning to think in Clojure.

Worth your attention:
• Eric Normand Courses: https://ericnormand.podia.com
• Freshcode Clojure team: https://www.freshcodeit.com/services/...

Chapters:
00:00 - Watch in this episode
01:01 - Episode 11 introduction
02:00 - Jeremiah journey in Clojure
03:15 - Finding the first job in Clojure
04:21 - Clojure in the New York Times
06:50 - Clojure adoption and internal status
10:20 - Official introduction of Clojure into the New York Times
11:41 - Unexpected benefits and drawbacks of Clojure
14:19 - Team dynamics and onboarding
20:06 - Code clarity and complexity management
23:29 - Performance of Clojure in high traffic
27:16 - Clojure longevity
31:22 - Team collaboration
34:18 - The big question. Would you do it again?
38:52 - Questions from Cam Saul from Metabase