This repository offers code and references for a postgraduate course on Applied Software Engineering with AI. It targets developers and researchers aiming to integrate AI into software development, covering machine learning, LLMs, and AI-driven automation through practical examples and curated links. The resources bridge theoretical AI concepts with tangible applications in software engineering.

How It Works

The project is organized thematically across modules, detailing AI fundamentals, deep learning (e.g., recommendation systems), and Web AI using TensorFlow.js. It explores prompt engineering, AI agents (MCPs) for automation, and deploying AI models locally or via APIs. Practical implementation is emphasized through numerous code examples, interactive demos, and tutorials.

Quick Start & Requirements

This is a collection of code examples, not a single application. Setup involves cloning individual examples and installing their dependencies (Node.js, Python, TensorFlow.js). A known issue exists with @tensorflow/tfjs-node installation on Windows. Links to external docs, demos, and resources are provided per module.

Highlighted Details

• Broad AI applications in software engineering, including RAG and semantic search.
• Focus on in-browser AI development with TensorFlow.js.
• Practical exploration of AI agents, MCPs, and automation tools.
• Guidance on local models (Ollama) and orchestrated APIs (OpenRouter).
• Examples of RAG implementation using JavaScript and Neo4j.

Maintenance & Community

Maintained by unipds-engenharia-de-ia-aplicada. No specific community channels or roadmap are detailed. Mentions of developers like ErickWendel and nico-martin suggest active community engagement.

Licensing & Compatibility

The provided README does not specify a software license, leaving terms of use, distribution, and commercial compatibility unclear.

Limitations & Caveats

A known limitation is the failure to install @tensorflow/tfjs-node on Windows. As a collection of examples, users must assess individual component complexity and compatibility. The lack of licensing information is a significant adoption caveat.