Motivation:
The motivation for this project was to make available to novice students (first-year computer science students) topics on software development models and the criteria by which one model or another could be selected based on the project's characteristics and the team developing it. It was later seen that it could be a valuable tool for advanced courses and even software developers.

Purpose:
This study describes the development of SoftWise, an AI-powered chatbot designed to assist students in Software Engineering courses and IT project development. SoftWise helps students learn about development models, from classical to agile. It also assists in selecting the most appropriate software development model based on key project criteria. It also provides bibliographical references, linkographies, and other relevant documentation.

Method and Tools:
An agile development approach was followed, extracting requirements from domain experts and learning outcomes from Software Engineering and Software Engineering Workshop courses. The chatbot workflow was designed in BotPress. The knowledge base was compiled from related course bibliographies and web-based information sources accessible through available Large Language Models (LLMs) such as GPT-4 8B, Google's Gemini (6B), Meta's Llama 2 (8B), and DeepSeek-V from DeepSeek AI Co.

Results: SoftWise
AI-powered assistant is currently at Technology Readiness Level 3 (TRL 3), available at https://cdn.botpress.cloud/webchat/v3.0/shareable.html?configUrl=https://files.bpcontent.cloud/2025/02/01/00/20250201002502-DK27PLXJ.json
After being tested in a controlled lab environment, it is currently being evaluated by experts using the Software Usability Scale (SUS). The next phase will involve students from related courses, which will allow the chatbot to be validated and its effectiveness as an educational tool to be measured.

Keywords: AI-powered assistant, chatbot, Software Engineering learning assistant, LLMs.