The Em Dash – Virtual Edition is part of Baton Rouge’s applied AI engineering meetup series, hosted by Obney.ai in partnership with the LSU Division of Computer Science & Engineering. Normally we gather in person on LSU’s campus—but this month, we’re taking it online to feature a world-class guest.
This session explores how pairing Large Language Models (LLMs) with REPL-driven development can transform them from brittle code generators into reliable engineering partners.
You’ll learn:
- Why REPLs matter for interactive, iterative AI workflows
- How they address LLM weaknesses like context loss and hallucination
- Why Clojure’s “code as data” model makes it an ideal fit for agentic AI development
- How specs and generative testing push LLMs beyond autocomplete into real engineering
👨💻 Speaker: Joe R. Smith, Co-founder/CEO of Doomsun (formerly Cognitect, Nubank, Crescent) — with 20+ years in distributed systems, decision engines, and functional programming.
Expect hard truths, practical insights, and a live demo. Whether you’re a researcher, engineer, or AI enthusiast, you’ll walk away with a clearer sense of what it takes to make LLMs actually useful.
