Transformer

A vector database is a specialized database optimized for storing, indexing, and querying high-dimensional vector embeddings. They enable fast similarity search, which is critical for RAG systems, recommendation engines, and semantic search applications. Popular vector databases include Pinecone, Weaviate, Qdrant, and pgvector for PostgreSQL.

A context window is the maximum amount of text (measured in tokens) that an LLM can process in a single interaction, encompassing both the input prompt and the generated output. Larger context windows allow models to handle longer documents, maintain extended conversations, and reason over more information at once. Context window sizes have grown rapidly — from 4K tokens in early GPT models to over 1M tokens in current models like Claude.

An embedding is a dense numerical vector representation of data — such as text, images, or code — in a high-dimensional space where semantically similar items are positioned closer together. Embeddings are fundamental to semantic search, recommendation systems, and RAG pipelines. They are generated by specialized models and typically stored in vector databases for efficient similarity lookups.

Fine-tuning is the process of further training a pre-trained AI model on a smaller, domain-specific dataset to adapt it for a particular task. Instead of training from scratch, fine-tuning adjusts existing model weights, which is significantly cheaper and faster. Common approaches include full fine-tuning, LoRA (Low-Rank Adaptation), and instruction tuning for aligning model behavior with specific requirements.

ETL (Extract, Transform, Load) is an automated data processing pattern where data is extracted from source systems, transformed into a desired format or structure, and loaded into a target system like a data warehouse. Modern variations include ELT, where raw data is loaded first and transformed in place. ETL pipelines are essential for automating data integration, reporting, and feeding clean data into ML training workflows.

Reinforcement learning (RL) is a machine learning paradigm where an agent learns optimal behavior by interacting with an environment and receiving rewards or penalties. RLHF (Reinforcement Learning from Human Feedback) is a key technique used to align LLMs with human preferences, making their outputs more helpful and safe. RL is also behind breakthroughs in game-playing AI and robotics.