Fintech & AI Glossary

A large language model (LLM) is a type of artificial intelligence trained on vast text data to understand, generate, and…

Retrieval-Augmented Generation (RAG) is an AI architecture that combines information retrieval with text generation. Ins…

A vector database is a specialized database system designed to store, index, and query vector embeddings — numerical rep…

An embedding is a numerical vector representation of data — text, images, or other content — generated by an AI model. E…

AI inference is the process of running a trained machine learning model on new input data to produce predictions, classi…

A context window is the maximum number of tokens — both input and output — that an AI model can process in a single inte…

Fine-tuning is a machine learning technique where a pre-trained model is further trained on a smaller, task-specific dat…

Prompt engineering is the practice of designing, refining, and optimizing input prompts to AI models to produce desired …

AI hallucination is a phenomenon where an AI model generates content that is factually incorrect, nonsensical, or not gr…

Anti-Money Laundering (AML) refers to the comprehensive set of laws, regulations, and procedures designed to prevent cri…

Know Your Customer (KYC) is the mandatory process by which financial institutions verify the identity of their clients a…

AI data residency refers to the requirement that data used for AI processing — including training data, inference inputs…

Semantic search is an AI-powered search technique that understands the meaning and intent behind search queries rather t…

Tokens are the fundamental units of text that AI language models process. Text is broken into tokens — which can be word…

Data governance is the comprehensive framework of policies, processes, roles, and technologies that ensures an organizat…

The General Data Protection Regulation (GDPR) is the European Union's comprehensive data privacy and security law that g…

Chunking is the process of dividing large documents into smaller, semantically meaningful segments before embedding and …

ETL (Extract, Transform, Load) is a data integration process that extracts data from source systems, transforms it into …

Model Risk Management (MRM) is the comprehensive framework of policies, procedures, and controls that financial institut…

A data warehouse is a centralized repository designed to store, manage, and analyze large volumes of structured data fro…

Transformer architecture is a deep learning model design introduced in the 2017 paper 'Attention Is All You Need.' Unlik…

The attention mechanism is a neural network component that enables models to dynamically focus on the most relevant part…

A neural network is a computing system inspired by biological neural networks in the human brain. It consists of layers …

Deep learning is a subset of machine learning that uses neural networks with multiple hidden layers (deep neural network…

Machine learning (ML) is a subset of artificial intelligence that enables systems to automatically learn and improve fro…

Few-shot learning is a machine learning technique where a model performs a new task after being shown only a small numbe…

Zero-shot learning refers to an AI model's ability to perform tasks it was never explicitly trained on, without being sh…

Chain-of-thought (CoT) reasoning is a prompting technique that elicits step-by-step reasoning from AI models before they…

An AI agent is an autonomous software system that perceives its environment, makes decisions, and takes actions to achie…

An MCP Server implements the Model Context Protocol, an open standard that enables AI models to securely access external…

AI orchestration is the coordination and management of multiple AI components, tools, and services to execute complex wo…

Temperature is a parameter in AI language models that controls the randomness of output generation. It ranges from 0 to …

Tokenization in AI is the process of breaking text into smaller units called tokens — typically words, subwords, or char…

Grounding in AI refers to the practice of basing model outputs on verifiable, external sources rather than relying solel…

RAG evaluation is the process of measuring the quality and effectiveness of retrieval-augmented generation systems. It a…

Multimodal AI refers to AI models that can process and generate multiple types of data — text, images, audio, video, and…

Constitutional AI is an AI training and alignment approach developed by Anthropic that uses a set of principles (a const…

An open-source LLM is a large language model whose weights, architecture, and often training code are publicly available…

On-premise AI deployment refers to installing and running AI models on an organization's own hardware and infrastructure…

A reasoning model is an AI model optimized for multi-step logical reasoning and problem-solving, typically using chain-o…

A data lake is a centralized repository that stores vast amounts of raw data in its native format until needed. Unlike d…

A lakehouse is a modern data architecture that combines the flexibility of a data lake with the reliability and performa…

A data pipeline is a set of automated processes that move and transform data from source systems to destination systems …

Streaming data is data that is generated continuously by multiple sources and sent incrementally for processing in real-…

Data lineage is the process of tracking and visualizing the flow of data as it moves from source systems through transfo…

A data catalog is a metadata management tool that creates an inventory of all data assets within an organization, making…

Cosine similarity is a mathematical metric used to measure the similarity between two vectors in a multi-dimensional spa…

Approximate Nearest Neighbor (ANN) is a search algorithm that finds items in a high-dimensional space that are approxima…

HNSW (Hierarchical Navigable Small World) is a state-of-the-art algorithm for approximate nearest neighbor search that b…

OLAP (Online Analytical Processing) is a computing approach that enables fast, interactive analysis of multidimensional …

Basel III is a comprehensive set of international banking regulations developed by the Basel Committee on Banking Superv…

The EU AI Act is a landmark European Union regulation that establishes a comprehensive legal framework for artificial in…

PCI DSS (Payment Card Industry Data Security Standard) is a set of security standards designed to ensure that all compan…

SOC 2 is a framework for auditing and reporting on service organizations' controls related to security, availability, pr…

DORA (Digital Operational Resilience Act) is a European Union regulation that establishes a comprehensive framework for …

RegTech, short for Regulatory Technology, refers to the use of technology, particularly AI and machine learning, to stre…

SR 11-7 is a supervisory guidance issued by the Federal Reserve and the Office of the Comptroller of the Currency (OCC) …

The Financial Action Task Force (FATF) is an intergovernmental organization that sets international standards for combat…

Explainable AI (XAI) refers to methods and techniques in artificial intelligence that make the outputs and decisions of …

AI governance is the framework of policies, processes, and controls that organizations establish to ensure their AI syst…

A robo-advisor is a digital platform that provides automated, algorithm-driven financial planning and investment managem…

Algorithmic trading AI refers to the use of artificial intelligence and machine learning algorithms to execute trades in…

Credit scoring AI is the application of machine learning and artificial intelligence to assess the creditworthiness of b…

Fraud detection AI is the application of machine learning and artificial intelligence technologies to identify and preve…

Financial sentiment analysis is the application of natural language processing and machine learning to quantify and inte…

Alternative data refers to non-traditional data sources used by financial institutions, hedge funds, and investors to ga…

Natural Language Processing (NLP) is a branch of artificial intelligence that enables computers to understand, interpret…

Document intelligence is the application of AI technologies, including computer vision, natural language processing, and…

WealthTech is a category of financial technology that focuses on digital solutions for wealth management, investment adv…

RegTech AI is the application of artificial intelligence and machine learning technologies to streamline and enhance reg…

Master Data Management (MDM) is a comprehensive discipline and technology framework that ensures the consistency, accura…

Data quality is a measure of the condition of data based on factors such as accuracy, completeness, consistency, timelin…

Business Intelligence (BI) is a technology-driven process for analyzing data and presenting actionable information to he…

API integration is the process of connecting different software applications and systems through Application Programming…

Open banking is a regulatory and technological framework that requires banks to share customer financial data with autho…

A feature store is a centralized data management platform that enables machine learning teams to define, discover, store…

Data mesh is a decentralized data architecture paradigm that shifts ownership of data from a central data team to domain…

Real-time data processing is the continuous ingestion, processing, and analysis of data as it is generated, with minimal…

Change Data Capture (CDC) is a data integration pattern that identifies and captures changes made to data in a source da…

Columnar storage is a data storage format that organizes data by columns rather than rows, enabling efficient compressio…

MiFID II (Markets in Financial Instruments Directive II) is a comprehensive European Union regulatory framework that gov…

Transaction monitoring is the process of continuously analyzing customer transactions to detect suspicious activity that…

A Suspicious Activity Report (SAR) is a confidential document that financial institutions and other regulated entities m…

Know Your Business (KYB) is a due diligence process that financial institutions and regulated entities use to verify the…

Customer Due Diligence (CDD) is the process of identifying and verifying the identity of customers, assessing their risk…

Stress testing is a risk management technique used by financial institutions to evaluate how their portfolios, balance s…

IFRS 9 is an International Financial Reporting Standard that governs the accounting treatment of financial instruments, …

The Sarbanes-Oxley Act of 2002 (SOX) is a landmark US federal law enacted in response to major corporate accounting scan…

An Open Banking API is a standardized application programming interface that enables third-party financial service provi…

Crypto compliance refers to the set of regulatory obligations, policies, and procedures that cryptocurrency businesses, …

AI latency is the time delay between sending a request to an AI model and receiving the complete response. In the contex…

AI throughput is the rate at which an AI inference system processes requests, typically measured in tokens per second or…

Model quantization is a technique that reduces the precision of numerical values in a neural network model to decrease i…

A Graphics Processing Unit (GPU) is a specialized processor originally designed for rendering graphics that has become e…

An AI benchmark is a standardized evaluation framework that measures and compares the performance of AI models across sp…

Reinforcement Learning from Human Feedback (RLHF) is a machine learning technique that fine-tunes AI models using human …

Model parameters are the numerical weights and biases that define the behavior of a machine learning model. In the conte…

Hybrid search is a retrieval technique that combines multiple search methods, typically semantic search using vector emb…

Reranking is a two-stage retrieval technique used in information retrieval and RAG systems that first retrieves a broad …

Agentic AI refers to artificial intelligence systems that can autonomously pursue goals, make decisions, and take action…

Knowledge distillation is a model compression technique where a smaller, simpler model (the student) is trained to repli…

A RAG pipeline is the end-to-end system architecture that implements Retrieval-Augmented Generation, encompassing all co…

A document loader is a component in a RAG pipeline that ingests documents from various sources and formats, parsing them…

A vector store is a specialized database or storage system designed to efficiently store and retrieve vector embeddings …

Dense retrieval is a search approach that uses dense vector embeddings — continuous numerical representations generated …

Sparse retrieval is a search approach that uses sparse vector representations — high-dimensional vectors where most dime…

Query expansion is a technique in information retrieval that enhances the original user query by adding related terms, s…

Context compression is a technique in RAG systems that reduces the amount of retrieved context passed to the LLM while p…

A knowledge graph is a structured representation of knowledge that organizes information as entities (nodes) and their r…

A multi-agent system is an AI architecture where multiple autonomous agents collaborate to accomplish complex tasks that…

Tool use is a capability of AI agents that enables them to interact with external systems, APIs, and data sources to per…

Function calling is a capability of large language models that allows them to detect when a user's request requires invo…

A system prompt is a set of instructions provided to an LLM at the beginning of a conversation that defines the model's …

Guardrails are safety mechanisms and constraints implemented in AI systems to prevent harmful, inappropriate, or non-com…

A model card is a standardized documentation framework that provides transparent, structured information about an AI mod…

AI safety is the field of research and practice focused on ensuring that AI systems operate reliably, ethically, and ben…

AI alignment is the field of research focused on ensuring that AI systems behave in accordance with human values, intent…

In-context learning is a capability of large language models where they learn to perform new tasks from examples provide…

Emergent behavior in AI refers to capabilities or behaviors that arise spontaneously in large-scale models as a result o…

Optical Character Recognition (OCR) is a technology that converts different types of documents — such as scanned paper d…

InsurTech is the term used to describe the application of technology innovations — particularly artificial intelligence,…

LendTech refers to the application of technology — particularly artificial intelligence, machine learning, alternative d…

PayTech refers to the technology and innovations that enable, process, and secure digital payments and transactions betw…

Embedded finance is the integration of financial services — including payments, lending, insurance, banking, and investm…

ESG AI scoring is the application of artificial intelligence — particularly natural language processing, machine learnin…

Portfolio Optimization AI refers to the application of artificial intelligence techniques — including machine learning, …

Trade Surveillance AI refers to the application of artificial intelligence — particularly machine learning, natural lang…

Market Surveillance AI refers to the application of artificial intelligence to monitor and analyze entire financial mark…

Conversational AI for Finance refers to the application of artificial intelligence technologies — particularly natural l…

AI underwriting is the application of artificial intelligence — particularly machine learning, deep learning, and predic…

Predictive analytics in finance refers to the use of statistical models, machine learning algorithms, and data mining te…

Contract intelligence, also known as contract analytics or AI contract review, is the application of artificial intellig…

A digital twin in finance is a virtual replica of a financial system, process, portfolio, or institution that uses real-…

Risk Model AI refers to the application of artificial intelligence — particularly machine learning, deep learning, and s…

Shadow AI in finance, also known as stealth AI or unauthorized AI, refers to the use of artificial intelligence tools, p…

Open finance is an evolution of open banking that extends the concept of customer-permissioned data sharing beyond payme…

SupTech, short for Supervisory Technology, refers to the use of innovative technology — particularly artificial intellig…

Fintech, short for financial technology, refers to the use of technology — particularly digital platforms, artificial in…

Blockchain analytics is the application of data analysis techniques — including AI, machine learning, graph analysis, an…

AI procurement in finance refers to the process of selecting, acquiring, and implementing artificial intelligence soluti…

Delta Lake is an open-source storage layer that brings ACID transactions to data lakes. Built on top of Apache Spark and…

Apache Iceberg is an open table format for huge analytic datasets. It is designed to manage petabyte-scale tables in dat…

Apache Parquet is an open-source columnar storage format designed for efficient data compression and encoding. Unlike ro…

Event sourcing is an architectural pattern that stores state changes as an immutable sequence of events rather than stor…

CQRS is an architectural pattern that separates read and write operations into different models. Commands handle write o…

A REST API (Representational State Transfer) is an architectural style for designing networked applications that use HTT…

GraphQL is a query language and runtime for APIs that enables clients to request exactly the data they need. Unlike REST…

Pub/Sub (publish-subscribe) is a messaging pattern where senders (publishers) broadcast messages without knowing which r…

A message queue is a communication mechanism that enables asynchronous data exchange between distributed systems using a…

A schema registry is a centralized repository for managing and validating data schemas in streaming and event-driven arc…

Data observability is an organizational practice and technical capability that provides end-to-end visibility into the h…

A data contract is a formal agreement between data producers and data consumers that defines the structure, semantics, q…

Metadata management is the practice of collecting, organizing, governing, and leveraging metadata about an organization'…

OLTP (Online Transaction Processing) is a class of database systems designed to handle high volumes of short, atomic tra…

Data partitioning is a database optimization technique that divides large tables into smaller, more manageable segments …

Data clustering is a database optimization technique that organizes data within each partition based on the values of sp…

A materialized view is a database object that stores the results of a query as a physical table, which is periodically r…

A dbt model is a SQL-based data transformation that defines how raw data is transformed into analytics-ready tables, vie…

Query optimization is the practice of improving the performance of database queries through techniques such as partition…

Data fabric is an architectural approach that connects disparate data sources across on-premises and cloud environments …

BCBS 239 is a regulatory framework issued by the Basel Committee on Banking Supervision that sets principles for effecti…

FRTB is a Basel Committee regulatory framework that fundamentally revises how banks calculate market risk capital requir…

A Legal Entity Identifier (LEI) is a unique 20-character alphanumeric code that identifies legally distinct entities par…

ISO 20022 is a global standard for financial messaging that provides a common language for payment transactions, securit…

XBRL is an open international standard for digital business reporting that enables the electronic exchange of financial …

SWIFT GPI (Global Payments Innovation) is a SWIFT initiative that transforms cross-border payments by enabling real-time…

SEPA is a European Union payment integration initiative that harmonises euro-denominated payments across 36 European cou…

PSD2 is a European Union directive that regulates payment services and payment service providers across the EU and EEA. …

PSD3 is the European Union directive that updates and replaces PSD2, strengthening the regulatory framework for payment …

CSRD is a European Union directive that mandates comprehensive sustainability reporting for large companies and listed S…

SFDR is a European Union regulation that requires financial market participants and financial advisers to disclose how t…

Third-Party Risk Management is the process of identifying, assessing, monitoring, and mitigating risks associated with o…

A regulatory sandbox is a framework established by financial regulators that allows FinTech companies and financial inst…

CCPA is a California state law that enhances privacy rights and consumer protection for residents of California, United …

Cyber resilience in finance is the ability of financial institutions to prepare for, respond to, and recover from cyber …

Capital adequacy is a regulatory requirement that banks and financial institutions maintain minimum capital levels relat…

Liquidity risk is the risk that a financial institution cannot meet its short-term financial obligations as they come du…

Model validation is the process of evaluating and testing financial models to ensure they are conceptually sound, accura…

Backtesting is a statistical methodology that evaluates the accuracy and reliability of financial models by comparing th…

An audit trail is a chronological record of events, actions, and changes that provides documented evidence of the sequen…

Anomaly detection is an AI technique that identifies unusual patterns, outliers, or suspicious events in data that devia…

Time series forecasting is a machine learning technique that uses historical data points ordered by time to predict futu…

A Graph Neural Network (GNN) is a deep learning architecture designed to process data structured as graphs, where entiti…

Federated learning is a machine learning technique that trains AI models across decentralized data sources without centr…

Synthetic data in finance refers to artificially generated datasets that mimic the statistical properties of real financ…

Transfer learning is a machine learning technique where a model developed for one task is reused as the starting point f…

AutoML automates the end-to-end process of applying machine learning to real-world problems. It automates data preproces…

Model drift detection is the process of monitoring machine learning models in production to identify when their performa…

A/B testing is a statistical method that compares two versions of a system, model, or process to determine which perform…

Causal AI is a branch of artificial intelligence that goes beyond correlation to understand cause-and-effect relationshi…

Natural Language Generation (NLG) is a subset of artificial intelligence that automatically produces human-readable text…

Voice AI refers to artificial intelligence systems that understand, process, and generate human speech. In banking, Voic…

Computer vision is a field of artificial intelligence that enables machines to interpret and understand visual informati…

An NLP pipeline is a sequence of processing steps that transforms raw text into structured, analyzable data for natural …

Reinforcement learning is a machine learning paradigm where an agent learns to make sequential decisions by interacting …

AI model monitoring is the practice of continuously tracking production machine learning models to ensure they perform a…

Data augmentation is a technique that increases the diversity of training data without collecting new data. In financial…

Knowledge distillation is a machine learning technique where a smaller, simpler model is trained to replicate the behavi…

Multimodal AI refers to artificial intelligence systems that process and integrate multiple types of data including text…

Privacy-preserving AI encompasses techniques that enable machine learning while protecting sensitive data. Key approache…

Model compression refers to a collection of techniques designed to reduce the size, memory footprint, and computational …

Prompt injection is a security vulnerability and attack vector targeting large language models in which an adversary cra…

Vector similarity is a mathematical measure of how close or related two vector embeddings are in a high-dimensional spac…

Text-to-SQL is a natural language processing technology that converts human language questions into structured SQL queri…

Agent memory is a critical architectural component of AI agent systems that enables agents to retain, recall, and utiliz…

Structured output in AI refers to the capability of large language models to generate responses that conform to a predef…

Context length, also referred to as the context window, is the maximum number of tokens that a large language model can …

Mixture of Experts is a neural network architecture that divides the model into multiple specialized sub-networks called…

Multimodal embedding is a technique that generates unified vector representations of data from different modalities, suc…

AI hallucination mitigation encompasses the set of techniques, methodologies, and architectural approaches designed to r…

Financial forecasting AI refers to the application of artificial intelligence, particularly machine learning and deep le…

Earnings analysis AI refers to the application of artificial intelligence, particularly natural language processing, mac…

Due diligence AI refers to the application of artificial intelligence technologies, including document intelligence, nat…

Financial modeling AI refers to the application of artificial intelligence, machine learning, and natural language gener…

KYC automation refers to the use of artificial intelligence, machine learning, optical character recognition, and docume…

AML automation refers to the application of artificial intelligence and machine learning technologies to automate and en…

Wealth management AI refers to the application of artificial intelligence technologies, including machine learning, natu…

Invoice processing AI refers to the application of artificial intelligence technologies, including document intelligence…

Credit risk AI refers to the application of artificial intelligence and machine learning technologies to assess, measure…

Regulatory reporting AI refers to the application of artificial intelligence technologies, including natural language ge…

SAMA Regulations refer to the comprehensive regulatory framework established by the Saudi Central Bank, formerly known a…

DIFC Financial Regulations refer to the comprehensive legal and regulatory framework established by the Dubai Internatio…

MAS Regulations refer to the comprehensive regulatory framework established by the Monetary Authority of Singapore (MAS)…

Market risk refers to the potential for financial losses arising from adverse movements in market prices, including inte…

Value at Risk (VaR) is a statistical measure used by financial institutions to quantify the potential loss in value of a…

A financial data platform is a comprehensive technology infrastructure that integrates, manages, analyzes, and delivers …

A market data feed is a continuous stream of real-time financial market data transmitted from exchanges, trading venues,…

Data warehouse modernization in financial services refers to the strategic process of migrating and transforming legacy …

Unstructured data in finance refers to the vast quantities of information that financial institutions generate and colle…

Real-time analytics in financial services refers to the practice of analyzing financial data as it is generated and inge…

LoRA (Low-Rank Adaptation) is a parameter-efficient fine-tuning technique that reduces the number of trainable parameter…

QLoRA (Quantized Low-Rank Adaptation) is a fine-tuning technique that combines 4-bit quantization of the base model with…

Parameter-Efficient Fine-Tuning (PEFT) is a family of techniques that adapt large pre-trained models to specific tasks b…

Instruction tuning is a fine-tuning technique where a language model is trained on pairs of instructions and desired res…

Catastrophic forgetting is a phenomenon in machine learning where a model loses previously learned knowledge when traine…

Training data for financial AI refers to the curated datasets used to train, fine-tune, or adapt AI models for financial…

A model card is a standardized document that provides transparent, structured information about an AI model's intended u…

A base model, also known as a foundation model, is a large AI model pre-trained on vast and diverse datasets that serves…