Beyond Prompt Engineering: PARLA AI and the Ability to Choose the Right Model for Each Business Task
- LLMs
Date
Wed 14 May
Time
16:00 - 16:15 CEST
Location
Tech Stage
About this session
Giada Franceschini explores advanced strategies to optimize the use of language models in the business world, going beyond the limitations of simple prompt engineering. Through a comparative analysis of the main LLMs and their distinctive features, the talk will present customization techniques such as Fine-tuning, LoRA, RAG, and Distillation. The presentation highlights the complexity of choosing the right model in a constantly evolving landscape and culminates in the demonstration of the PARLA AI solution—the first Italian LLM aggregator that implements an intelligent router to automatically select the most suitable model for each task, optimizing efficiency, accuracy, and sustainability for Italian companies.

