The field of artificial intelligence (AI) is undergoing significant changes, with enterprises integrating a range of new technologies. As we navigate 2023, apparently the year AI becomes obtainable by virtually every business, it’s evident that the AI stack is expanding beyond just algorithms and data. This talk will provide a straightforward overview of the new infrastructure that traditional DevOps should understand in the the emerging domain of AIOps.
Key Topics:
Data Infrastructure: A look at the role of data lakes, warehouses, and the need for data integration and ETL tools to train or fine-tune large language models
Data Labeling: An exploration of the methods and importance of both manual and automated data labeling processes.
Vector Databases: An introduction to vector databases and their application in handling high-dimensional data.
Foundational Models: A discussion on transfer learning, highlighting models that are readily available for use from platforms like Hugging Face.
Fine-Tuning vs. Training: A comparison of the processes of fine-tuning existing models and training new models from the ground up.
LLMS & Multi-Modal Models: An examination of language model-based learning systems and the function of multi-modal models.
Transition to AIOps: A study on how AI operations (AIOps) will affect the DevOps field, with a focus on AI-driven tools and practices.
Ethical AI and Security Considerations: A review of current concerns in AI ethics, including bias detection, model explainability, and security measures.
This talk aims to provide a clear and concise overview of the evolving AI technologies that enterprises are likely to adopt by 2023.