# Artificial Intelligence and Why It Matters in Banking **AI Glossary** Every AI conversation in banking brings new vocabulary. This glossary explains common AI terms, what it means for banks, and how nCino applies it to help financial institutions operate faster, smarter and with less risk. A plain-language glossary of AI terms in banking—from LLMs to agentic AI—and how nCino applies each one. Understand the vocabulary behind AI banking. --- ## Featured Stats Gradient Cards - **2,700+ Around the world** — CUSTOMERS - **14+ Real credit decisions and workflow data trains every model behind these terms. ** — Years of Banking Context - **60-70% Faster relationship reviews with Analyst Digital Partner as reported by our customers** — FASTER RELATIONSHIP REVIEWS --- ## General AI Terms The underlying technology that applies across industries. ### Artificial intelligence \(AI\) In plain language: An engineered or machine-based system that can, for a given set of objectives, generate outputs such as predictions, recommendations, or decisions influencing real or virtual environments.¹ How nCino uses it: At nCino, AI means the AI that helps bankers work faster and improves the platform over time. AI is not a feature we add to an existing product. AI is the foundation of the next generation of the platform. ### Large language model \(LLM\) In plain language: The technology that reads, understands and generates natural language. Leading models serve general purposes rather than banking specifically. LLMs typically also serve as foundation models — the general-purpose model that specialized applications are built off of. How nCino uses it: Banking Advisor and the Digital Partners use LLMs. The models power the experience, but without the right banking context they produce generic output. nCino supplies more than 14 years of banking-specific operational context, which makes every output specific to your institution. nCino uses LLMs that larger model providers manage and host, because those providers build the most advanced models available. nCino then builds banking applications on top of them. ### Retrieval-Augmented Generation \(RAG\) In plain language: A technique that gives AI access to institutional and specific information before it answers. Without it, an LLM gives a brilliant but generic answer. With it, the answer is grounded in your actual data. How nCino uses it: nCino uses RAG in Banking Advisor Knowledge Base to source chat responses based off your organization’s document repository. ### Predictive AI In plain language: AI that learns from historical outcomes to estimate what is likely to happen next, with a confidence level attached. It produces scores and forecasts rather than conversation. How nCino uses it: Predictive AI estimates credit risk — default likelihood, delinquency, shifting portfolio exposure; predicts events — when a loan will close, where a file will stall, where volume is heading; and turns both into recommendations about what to do next. Each model is purpose-built for a specific task and trained on nCino’s proprietary lending data. Predictions inform decisions; people make them. ### AI Agent In plain language: “A software program that can interact with its environment, collect data, and use that data to perform self-directed tasks that meet predetermined goals.”² How nCino uses it: \[Digital Partners\]\(https://www.ncino.com/news/ncino-introduces-digital-partners-role-based-agents\) are role-based agents purpose-built for specific banking functions. Each digital partner addresses distinct pain points and outcomes, executing workflows, tools and sub-agents, and providing recommendations within their defined scope \(see: Agentic AI\). ### Agentic AI In plain language: “An autonomous AI system that can act independently to achieve pre-determined goals. “Agentic” indicates agency — the ability of these systems to act independently, but in a goal-driven manner.”³ How nCino uses it: \[Digital Partners\]\(https://www.ncino.com/news/ncino-introduces-digital-partners-role-based-agents\) are role-based agents that operate across the banking workflow without needing a banker to prompt every step, within the boundaries the institution set. ### Agent Orchestration In plain language: “The process of coordinating multiple specialized AI agents within a unified system to efficiently achieve shared objectives. Rather than relying on a single, general-purpose AI solution, agent orchestration employs a network of AI agents, each designed for specific tasks, working together to automate complex workflows and processes.” How nCino uses it: nCino builds orchestration into the Agentic Operating System, so a complex banking process runs end to end rather than as a set of disconnected tasks. The orchestration layer coordinates the agents, tools, and data flows that the process needs, and your institution does not build that coordination logic itself. ### Human-in-the-loop \(HITL\) In plain language: A design principle ensuring a human approves before AI takes actions above a certain threshold of impact or risk. How nCino uses it: Human-in-the-loop controls are built into the agent platform by design, because regulators and chief risk officers require them before approving agent deployment. nCino typically defaults create/update/delete operations to require human oversight. This governance feature allows you to move at speed while staying inside your own risk appetite. ### Observability In plain language: The ability to understand what a system is doing in real time. It means the ability to watch agent decisions, flag deviations from expected patterns and produce records that satisfy audit and regulatory explanation. How nCino uses it: \[Operations Analytics\]\(https://www.ncino.com/operations-analytics\) is a feature that measures platform performance and process improvement. OA reports how long a workflow takes, where it stalls and whether you can explain a decision to a regulator. ### Tokens In plain language: AI tokens are the fundamental unit of AI work, measured in characters of text, pieces of an image, or slices of audio that an AI model can process. How nCino uses it: nCino measures consumption through Intelligence Units. They measure each time an agent action or capability runs, or a system makes an API call. Usage is counted on what is actually consumed, rather than estimated up front. IUs make intelligence usage visible and accountable alongside your platform subscription, and they cover third-party agent usage through the gateway. You see which capabilities your institution uses, who uses them, and at what volume.⁷ ### Conversational AI In plain language: Conversational artificial intelligence \(AI\) is "a technology that allows software to process and respond to human voice or text-based conversations."⁸ How nCino uses it: \[Banking Advisor\]\(https://www.ncino.com/banking-advisor\) and \[Mortgage Advisor\]\(https://www.ncino.com/mortgage/us/mortgage-pos\) are nCino's unified conversational interfaces where users interact with AI through natural language. Bankers and loan officers simply ask questions or request actions rather than navigating complex system interfaces. --- ## Connectivity How systems, agents and data reach each other — and where all of it runs. ### Application Programming Interface \(API\) In plain language: A standardized way for software systems to communicate and exchange data reliably. APIs are the connection point between your core banking system and nCino. How nCino uses it: An open API-first architecture lets you connect nCino to any system in your stack and extend the platform beyond what nCino ships. The API layer also serves as the governed gateway for third-party systems and agents. ### Model Context Protocol \(MCP\) In plain language: A technical standard published in 2024 and now widely adopted, creating a common language for AI agents to access tools, data sources and capabilities from external systems.⁵ How nCino uses it: nCino develops MCP in two directions. nCino builds its own MCP servers so outside clients connect into nCino, and those servers today target the administrator for mortgage and global banking, and the banker for business banking. nCino also develops MCP support within the user-facing AOS experience, so an agent you build in nCino connects to your own MCP servers. ### Agent2Agent \(A2A\) In plain language: A 2025 standard allowing AI agents from different vendors to discover, coordinate and delegate work to each other without a human in the middle.⁶ How nCino uses it: A2A makes agent orchestration work across vendor boundaries. An agent elsewhere in your institution delegates a credit analysis subtask to the nCino Analyst Digital Partner and continues its own workflow. These interactions route through the same governed gateway, so nCino logs and controls cross-vendor autonomy rather than opens a direct data path. ### Integration Platform as a Service \(iPaaS\) In plain language: A cloud-based platform that provides tools, services and infrastructure to build, deploy and manage integrations between applications, data sources and systems without needing on-premises hardware or extensive technical resources. How nCino uses it: \[Integration Gateway\]\(https://www.ncino.com/solutions/integrations\) is the integration platform that is tailor-made for banks and credit unions to enable fast, secure, and scalable integration between cores, fintech apps, LOS, CRM, and more, without overwhelming internal IT teams. It empowers financial institutions to accelerate digital transformation, streamline operations, and deliver connected customer and member experiences. ### AWS In plain language: Amazon's cloud computing platform that provides computing power, storage, networking and AI services on demand. How nCino uses it: nCino is migrating global solutions to AWS for faster scaling, lower infrastructure burden and access to real-time AI inference that on-premises architectures cannot match. ### Cloud vs. on-premises In plain language: Cloud means software that runs on third-party servers accessed over the network. On-premises means software that runs on servers you own and maintain. Most major financial institutions run a hybrid. How nCino uses it: Cloud infrastructure unlocks real-time agent inference at scale. Moving core workflows to the cloud gives you access to modern AI capabilities today, without the burden of buying and maintaining that compute yourself. ### Agentic Operating System \(AOS\) In plain language: An emerging software architecture designed to run, manage and coordinate autonomous AI agents rather than traditional applications. How nCino uses it: The AOS provides banking-specific guardrails, connects workflows across the ecosystem and ensures every AI action is grounded in 14+ years of banking context with full governance and compliance. Unlike basic chatbots, the AOS lets multiple specialized AI agents collaborate, share memory, access tools, and complete multi-step tasks across different business systems without constant human intervention. This means you can deploy agents without the 12–18-month engineering time most institutions face when building this internally, so you can move faster while staying audit-ready. --- **Brochure** ### Digital Partners Overview Explore nCino’s role-specific agentic AI partners, built to get banking work done for your team. [Download PDF](https://assets.ctfassets.net/ze344jjyx19q/3hcb8XPED7ssetd2N7L9Fz/a33b37d879e4090b490bea6eda93f6a6/nCino_Digital_Partners_Brochure.pdf) **Solution** ### Data, AI, Analytics nCino's AI is built on one of the most comprehensive perspectives in financial technology, which allows us to deliver what no other partner can [Explore the Solution](https://www.ncino.com/solutions/data-ai-analytics) **Customer Stories** ### What Real AI Strategy Looks Like Inside a Bank At Zions Bancorporation, that strategy is specific enough to name, sequence, and staff, and it's being shaped by leaders who are just as focused on their people as they are on their technology. [Read the Article](https://www.ncino.com/blog/AI-strategy-zion-bancorporation) --- ## See Banking Intelligence in Action Bring these concepts into your own workflows with a walkthrough tailored to your institution. [Request a Demo](/demo) --- ### Foot Notes ¹ Organisation for Economic Co-operation and Development. (n.d.). _OECD AI Principles overview_. OECD.AI. Retrieved October 6, 2026, from [https://oecd.ai/en/ai-principles](https://oecd.ai/en/ai-principles) ² Amazon Web Services. (n.d.). _What are AI agents?_ Retrieved October 14, 2026, from [https://aws.amazon.com/what-is/ai-agents/](https://aws.amazon.com/what-is/ai-agents/) ³ Amazon Web Services. (n.d.-b). _What is agentic AI?_ Retrieved October 14, 2026, from [https://aws.amazon.com/what-is/agentic-ai/](https://aws.amazon.com/what-is/agentic-ai/) ⁴ Downie, A., & Finio, M. (2025, February 26). _What is AI agent orchestration?_ IBM. [https://www.ibm.com/think/topics/ai-agent-orchestration](https://www.ibm.com/think/topics/ai-agent-orchestration) ⁵ Model Context Protocol. (n.d.). _What is the Model Context Protocol (MCP)?_ Retrieved October 14, 2026, from [https://modelcontextprotocol.io/](https://modelcontextprotocol.io/) ⁶ The Linux Foundation. (2025, June 23). _Linux Foundation launches the Agent2Agent Protocol Project to enable secure, intelligent communication between AI agents_ [Press release]. [https://www.linuxfoundation.org/press/linux-foundation-launches-the-agent2agent-protocol-project-to-enable-secure-intelligent-communication-between-ai-agents](https://www.linuxfoundation.org/press/linux-foundation-launches-the-agent2agent-protocol-project-to-enable-secure-intelligent-communication-between-ai-agents) ⁷ Deloitte. (2026, January 11). _Navigate the economics of AI: How tokenomics is reshaping AI costs and ROI_. [https://www.deloitte.com/us/en/services/consulting/articles/how-to-navigate-economics-of-ai.html](https://www.deloitte.com/us/en/services/consulting/articles/how-to-navigate-economics-of-ai.html) ⁸ Amazon Web Services. (n.d.). _What is conversational AI?_ Retrieved October 21, 2026, from [https://aws.amazon.com/what-is/conversational-ai/](https://aws.amazon.com/what-is/conversational-ai/) --- [View sitemap](https://www.ncino.com/api/markdown/sitemap)