> Markdown version of [/events/world-congress-2025/sessions/854-llm-local-inference](https://www.wearedevelopers.com/events/world-congress-2025/sessions/854-llm-local-inference). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# LLM Local Inference: Tools, Techniques, and Insights


- **Date:** Friday, Jul 11, 2025
- **Time:** 13:55–14:05 (10 min)
- **Room:** Airstream 1
- **Event:** World Congress 2025
- **Tags:** ai, llms, open source, python

## Description

Over the past few months, I’ve had the chance to work on a research and development project exploring the feasibility of adopting solutions powered by Small and Large Language Models—with one unique twist: inference had to be performed at the edge.

This journey led me to scratch the surface of the vast, fascinating, and rapidly evolving world of LLMs, with a particular focus on inference servers. Along the way, I encountered a fair share of head-scratching moments and valuable insights that I’m excited to share with you.

In this talk, we’ll cover the key concepts behind LLM inference, untangle the tricky jargon, and give you a glimpse into the primary tools and solutions you can leverage if you ever need or want to explore LLM local inference.

## Speaker

### [Riccardo Zoncada](https://www.wearedevelopers.com/@riccardo-zoncada)

Software Engineer @ xtream


## Related talks at this congress

- [Self-Hosted LLMs: From Zero to Inference](https://www.wearedevelopers.com/events/world-congress-2025/sessions/625-self-hosted-llms) — Cedric Clyburn, Roberto Carratalá
- [Inside the Mind of an LLM](https://www.wearedevelopers.com/events/world-congress-2025/sessions/831-inside-the-mind-of) — Emanuele Fabbiani
- [Exploring LLMs across clouds](https://www.wearedevelopers.com/events/world-congress-2025/sessions/525-exploring-llms) — Tomislav Tipurić
- [ Model Compression Techniques for Efficient LLM Deployment](https://www.wearedevelopers.com/events/world-congress-2025/sessions/811-model-compression) — Harshita Seth, Lavinia Ghita, Sergio Perez, Ziv Ilan