Generative AI Engineering
(field of studies: Computer Science)

• Programme: from deep learning to deployment
• Specialisation: courses and projects
• Labs: project-based, engineering-focused labs
• After graduation: AI and ML engineering roles

Generative AI Engineering is a specialisation within the Computer Science programme, a full-time Bachelor's course taught in English at the Polish-Japanese Academy of Information Technology (PJAIT). It is aimed at those who want to build, tune and deploy modern generative and multimodal AI systems.

Programme: from deep learning to deployment

The specialisation follows the whole lifecycle of an AI system, and you work with:

• deep learning architectures such as CNNs, RNNs, autoencoders, GANs and Transformers
• generative and multimodal models, including Vision Transformers, diffusion models and CLIP
• Retrieval-Augmented Generation (RAG) systems and agent-based systems
• efficient fine-tuning with LoRA and adapters, and model compression through quantisation, distillation and pruning
• model alignment (RLHF, DPO) and reinforcement learning, including applications in robotics
• AI and MLOps engineering: APIs, Docker, CI/CD, MLflow and Kubernetes

Specialisation: courses and projects

The specialisation courses are Deep Learning in Python, which covers fundamental and advanced techniques, model implementation and projects, and Generative AI, which deals with modern generative and multimodal models, optimisation, alignment and agent-based systems. Project-based classes follow: the first concentrates on AI and MLOps systems engineering and model-as-a-service deployment, the second on trends such as scaling laws, security and model economics, and the third prepares you for writing and presenting your thesis. In the sixth semester you are encouraged to take the elective Methods of Knowledge Engineering.

Labs: project-based, engineering-focused labs

The labs are project-based and engineering-focused. You implement models from scratch in NumPy, PyTorch and TensorFlow, build RAG, multimodal search and diffusion systems, and then turn them into working services with FastAPI, Docker, data pipelines, monitoring and deployments on Kubernetes.

After graduation: AI and ML engineering roles

Graduates are prepared to work as AI engineers, ML engineers or generative systems specialists, in fields such as IT, robotics, the creative sector, medicine and data analysis.

Because the programme stresses practical implementation, you can join both product teams and research and development teams. Thesis topics show the direction: a RAG system for knowledge retrieval, a multimodal image-text search based on CLIP, or the optimisation of language models with LoRA and quantisation.

Based on the programme pages at pja.edu.pl.