Glossary of Aisera Terms

Make sure you understand these terms and the have read the Prerequisites section before you proceed to build your applications/bots.

Aisera Products

Aisera Product	Description
Agent Assist	Agent Assist is an embedded application/bot that runs in your Agent Desktop such as ServiceNow, Salesforce, Zendesk, or Jira Service Desk and provides IT Helpdesk Agents with recommendations to resolve their cases or real time live agent conversations occurring over chat or phone. Currently supported on Zendesk, ServiceNow, and Salesforce Service Cloud and Jira Service Desk. Includes Agent Assist Analytics display page within the Aisera Admin UI. Terms not to use: Auto-resolve, TicketIQ, TicketAI, AI Assist Modules include: Case Summary: Quick summary of what’s happened in the Ticket thus-far. This helps tier 2 / tier 3 escalation resources quickly catch-up on what’s happened thus-far. Aisera Answer: RAG answer. Typically a richer answer than an end-user receives as the service desk often has access to more granular technical documentation than what’s available to the general employee/customer audience. Next Best Actions: Recommended next-steps based on learnings from historical similar cases. Field Predictions: Predicts how to categorize, assign, route, or prioritize based on historical ticket learnings. Can auto-apply predictions above a certain threshold (ie. above 90% confidence in prediction). Intelligent Swarming: Aisera offers intelligent swarming through analyzing various factors like ticket content, urgency, impact, and historical data to identify tickets that will likely require collaboration from multiple experts across various departments User Insight/Sentiment Analysis: Aisera’s Sentiment analysis uses natural language processing (NLP) and machine learning techniques to analyze and interpret the emotions expressed in text data from an utterance or the conversation thread associated with a user query. Case Wrap-Up: Allows IT technicians, engineers, or Customer Support Agents to look up the Case History for future issues or escalations. Case Creation: The Aisera platform includes out-of-the-box workflows for automated ticket creation and ticket management. Aisera can create and update cases in any ticketing system from any conversational channel, such as Slack, Microsoft Teams, or a custom conversational interface. Use the Ticket AI Agent Assist Channel. Add other channels, such as IVR as needed. See also: Overview of the Agent Assist App.
Aisera Assistant (formerly AI Copilot)	Aisera Assistant is an advanced, enterprise-grade employee copilot designed to enhance productivity and streamline workflows through proactive automation and intelligent task management. Leveraging the power of AiseraGPT and domain-specific large language models (LLMs), Aisera Assistant delivers personalized, context-aware support that caters to a wide range of business needs. With Aisera Assistant, users can upload documents locally and save previous conversations. Terms to Use: Agent, Universal Agent, Agent Orchestrator, Conversational AI Terms not to use: Bot, Copilot, AiseraGPT, App ConversationalAI 2.0: Flow Search See also: Overview of Aisera Assistant/Copilot
AI Ops Assistant/Agentic AI	This application can be built today using the AI Ops base application with an AI Copilot channel. Some of the capabilities are still in development. An AI system built to act as an autonomous agent(s). Capable of conversing, reasoning, and performing actions dynamically with minimal human intervention. Key Components of an Agentic AI: Autonomous: operates independently based on goals or prompts Multi-Fullfillment / Task Planning: able to dynamically orchestrate a diverse set of tools / skills / sub-agents (Knowledge Serving / RAG, Code Gen, Workflows, form and execute API requests, text-to-sql, etc.). Truly agentic agents should be able to invoke these tools sequentially and/or in parallel. They should be able to create a plan of actions and/or perform task decomposition. If the AI is limited to singular request/response tool calling, this is not agentic. Context-awareness: Ability to understand and maintain conversational context as well as adapt to environmental / user-specific context (environmental variability, user personalization / ACL adherence) Agentic agents need the ability to store/cache not just conversation history but also output/environment variables from previously invoked tools / sub-agents Proactivity: Can initiate tasks, not just respond to commands. See also: AI Ops Assistant
Universal Bot/Multi-Agent Agentic	An AI system in which multiple AI agents, each capable of autonomous task execution and reasoning, collaborate dynamically—either cooperatively or hierarchically—to fulfill complex, multi-step goals. Interoperability Choose Universal Bot instead of Domain Specific when building the bot.
Agentic AI for ITSM	Aisera's Agentic AI for ITSM represents a breakthrough in autonomous IT service management, delivering intelligent problem-solving capabilities that go beyond traditional scripted automation. Our agentic architecture enables autonomous decision-making, adaptive responses, and continuous learning to transform reactive IT support into predictive, autonomous service delivery. See also: Agentic AI for ITSM
AI Ops	The Aisera AIOps Platform is designed to help IT operations using AI that leverages AiseraLLM that's fine tuned for ITOps. This advanced solution offers a comprehensive approach to managing IT, DevOps, and Cloud operations, ensuring improved system uptime, proactive incident management, and enhanced operational efficiency. AI Ops includes: CMDB - Configuration Management Database Discovery - This is an option within the left menu of the Aisera Admin UI called AI Discovery - CMDB that allows you to let the AI determine and create a list of the Systems, Applications, Networks, Storage, and Business Services on your internal network inventory. See also: Inventory Transparency. Major Incidents - A critical part of the AIOps platform is the Major Incident Detection capabilities. By utilizing IT Operations-specific Aisera Large Language Model, Aisera AIOps provides an innovative approach to identifying and managing major incidents. See also: Major incident Identification and Configure an Aisera Major Incident. Intelligent Alerts - Intelligent Alerts allow you to set alerts on specific fields or on correlated fields, so you can see how one value that you're tracking affects another. See also: Financial Planning Operations. Incident (Ticket Clustering) - Ticket clustering leverages Aisera's Large Language Models (LLMs) and sophisticated techniques based on temporal and other ticket attributes to automatically group related support tickets. LLMs analyze the textual content of tickets, such as titles, descriptions, and comments, to identify underlying themes and similarities. See also: Incident (Ticket) Clustering. Ticket Clustering is also available in the Agent Assist application. See also: Overview of the AI Ops Application.
AI Search (Deprecated)	Aisera’s Enterprise AI Search offers users the ability to access content across multiple different external applications. Each team in a company utilizes multiple platforms (some related to ITSM include Servicenow, Google Suite, Sharepoint, Microsoft 360, Jira, Salesforce, and Zendesk) on a daily basis to create, update, and store knowledge. Knowledge is lost in these documents and becomes difficult to access when needed. Enterprise AI Search is a tool for users to search for knowledge across different platforms in the same place.
AI Voice	The AI Voice Bot allows you to analyze voice messages, interpret Intents, and respond with a fulfillment option. The Contact Center integration with the Aisera platform for Voice Assistant support is enhanced with updated micro services that orchestrate additional Speech-to-Text (STT) and Text-to-Speech (TTS) vendors. The AI Voice bot supports Voice Activity detection and updates to cleanup sessions on the Aisera platform after call termination. See also: AI Voice.

Aisera Terms

Aisera Term	Description
Agent	The term Agent can be used to refer to an application or bot, as well as to a human who is interacting with customers. In the Aisera Product Documentation, it is often used as Service Agent when referring to a human Agent and Application/Bot when referring to an Artificially Intelligent Agent.
AI Automation	The Aisera Administration Application contains a menu item called, AI Automation that includes tools to help you automate the flows (sequences of tasks your bot performs). Use this menu item to access the AI Workflow Studio, Event Studio, Hyperflow Studio, and Campaigns tools.
AI Lens	The AI Lens is a tool within the Aisera Admin UI that lets you simulate the interaction with your bot. The simulated input Requests are not captured within the Request Details page. AI Lens is similar to runtime analysis. See also: AI Lens
AI Discovery - CMDB	For future use.
AI Observability	This is a feature of the AI Ops application that allows you to view Major Incident and Intelligent Alerts.
Aisera Administration Application	The Aisera Gen AI Platform allows you to configure a myriad of settings for at both the tenant and the bot levels, using the Aisera Administration Application (also called the Aisera Admin UI). After you receive your tenant instance and login credentials, you can put your tenant URL into a browser window and log in to the Aisera Admin UI.
Aisera Admin UI	Short name for the Aisera Administration Application (see above).
AiseraGPT	The general designation for an Aisera bot (like ‘ChatGPT’), before it has a domain pack (content pack), channel or data source associated with it.
Aisera Gen AI Platform	The Aisera Gen AI Platform refers to the Aisera software in general. It represents the services and micro services running on the backend servers ‘in the cloud’ as well as the interactions that take place on a specific customer tenant or bot.
AI Workflow Studio	The AI Workflow Studio is a feature in the Admin UI that allows you to set the order of tasks for your application/bot to perform. You can use the AI Studio canvas to drag and drop nodes (no code or low code) to create a flowchart that your bot will follow when it runs. See also: AI Workflow Studio - and - Workflow Best Practices.
AI Workbench	The Aisera Administration Application contains a menu item called, AI Workbench that give you access to tools that help you optimize the entities (Requests, Intents, Tickets, and Knowledge Base Articles and Fulfillments) that are used within a flow. Use this menu item to access the tools that allow you to work on Unresolved Requests; Review Intents, Knowledge Articles, and Synonyms; Check the Health of your Intents; and Create Test Suites.
Analytics	The Aisera platform provides advanced Analytics capabilities, with pre-built reports & dashboards, as well as the ability to create custom reports & dashboards based on any available data in the platform. Some of the key metrics include resolution rate, escalation rate, abandoned rate, customer & employee satisfaction (CSAT/ESAT), MTTR, top unresolved intents, Knowledge Gaps, and top intents by sentiments. See also: Pre-Built Analytics.
APIs	Aisera provides APIs for conversation information, data ingestion, data source ingestion monitoring, and web socket connection. See also: APIs.
Audit Trail Logs	You can track all operations performed on your Aisera platform objects by choosing the Audit Trail option from the left navigation menu of the Aisera Administration Application. See also: Audit Trail Logs.
Benchmarking	The act of evaluating or checking something in comparison to a standard. Aisera's Large Language Model (LLM) benchmarking process was designed to evaluate the performance and capabilities of our AI-driven service automation solutions. The datasets used for the task evaluations were sampled from our Retrieval Augemented Generation (RAG) system in financial and legal & compliance domains. See also: Aisera LLM Benchmarking.
Bot	A Virtual Assistant or Agent.
Campaigns	Campaigns is a tools within the AI Automation feature that enables you to send broadcast messages to specific user groups through various communication channels (MS Teams, Slack, Glip, and custom SMTP). See also: Campaigns.
Channels	Channels are communications applications that your application/bot can integrate with and open from within. For instance, if you're already using an application like MS Teams or Slack, you can integrate your virtual assistant as a feature within those applications. You can also think of this as the ‘skin’ where your bot runs. See also: Channels.
Connector	This is a generic term used within Aisera to refer to Data Sources or Channels that are integrated with an Aisera tenant instance.
Content Generation	The term used to describe that your application/bot can generate Knowledge Base Articles. Use the Content Generation > Knowledge Generation menus in the Aisera Admin UI to initiate a Knowledge Generation Job. See also: Content Generation.
Conversation	A dialog between two or more parties. In Artificial Intelligence, the Conversation is the question and answer sequence between the application/bot user and the AI software.
Conversational AI	A technology that enables machines to engage in human-like dialogue through voice or text. It combines NLU/NLP, dialog management, and sometimes generative AI models to interpret, respond to, and learn from user input. Utterance: A single piece of user-provided input, typically a sentence or phrase. It represents what the user says or types at a given point in a conversation. Examples: “Hey I forgot my password” “Which medical plan has the lowest deductible?” Request: The question asked by the bot user, which is a structured representation of a user’s intent, often enriched with metadata. It typically includes the recognized intent, any extracted entities, and session context. It is what the system uses to take action or generate a response. In the context of Aisera, we consider a request to be a single transaction between Aisera and the user. For example: 1 Request: Utterance: “Hey what’s the wifi password” Response: “It’s aisera2025” 1 Request: Utterance: “Hey I need to unlock my account” Response: “Sure thing! First can you give me your Okta Verify Code” Utterance: “Yep it’s 87365” Response: “Your account has been unlocked!” Utterance: “Thanks, you’re a lifesaver!” Response: “No problem!” 2 Requests: Utterance: “Hey can you get me access to Acrobat and Illustrator” Response: “I’ve provisioned your account access to the adobe creative suite, which includes Acrobat and Illustrator” Utterance: “Awesome! Now can you draft an email to Tom letting him know I’ve got acrobat and run OCR on those scanned PDFs if he sends them over to me?” Response: “I’ve drafted your email below… Let me know if you’d like me to send it” Session: A single, ongoing conversation or interaction between the user and the AI, encompassing all exchanged utterances and context until the interaction ends. Sessions may persist across multiple turns / multiple different requests.
Conversational AI 1.0 (Still in Use)	This term refers to the original Aisera Gen AI Platforms traditional ICM search that is used to match Intents with Fulfillments by allowing the Aisera Admin to select a Fulfillment Engine search sequence for each application/bot. Customer Admins did not have access to this parameter.
Conversational AI 2.0 (Preferred Option)	Conversational AI 2.0 is the term that Aisera uses for it’s second product version that allows users to create bots that use LLM features instead of NLP features, and allows bot creators to use slot filling, Workflows, and Hyperflows to provide Fulfillment, instead of using the previous Fulfillment Engine. The Customer Administrator can designate an application/bot as a 2.0 bot by setting a toggle switch in the Advanced tab of the AiseraGPT Details window. See also: Creating a Conversational AI 2.0 Bot.
Data Model	The Aisera data model contains the fields for our internal vector database that must be mapped to an external data source before data can be used to answer Requests in an Aisera application/bot. The customer’s data source also has a data model, and the corresponding fields need to be mapped. See also: Field Mapping (Product Docs - you can share) See also: Mapping KB Article Fields to Generated Documents.
Data Source	The part of the connection between the Aisera Gen AI platform and the external Data Source that includes the data and the data mapping.
Disambiguation Questions	Follow-up questions to the original user Request that help to clarify the Intent (purpose) of the user’s Request. It is a best practice to limit the number of disambiguation questions and have the bot create a Ticket or escalate to a Live Agent to avoid frustrating the bot user.
Domain Specific LLMs	These LLMs are fine-tuned for use with specific tasks and domain knowledge. See also: Domain Specific LLMs
Event Studio	Aisera’s Event Studio empowers customers to build robust AI Automations triggered by a wider range of events, both internal and external. See also: Event Studio
Fulfillment	An answer that resolves the Request entered by an application/bot user.
Gen AI	A category of AI models that can generate new content, such as text, images, audio, video, or code, based on patterns learned from existing data.
Generative AI Learning	Allows you to run a Learning Job (System Job) against data that you have ingested. After you run the Learning Job, The Generative AI Learning window includes a Summary of the learning, and then breaks down the data into Ticket Learning, Knowledge Learning, and Case Insights window tabs. You can also use this feature with Conversation Logs.
Integration	The part of the connection between the Aisera Gen AI platform and the external Data Source that includes the Authorization and Authentication (Auth & Auth).
Intent (Still in Use)	The goal or purpose of a bot Request (what the user wants to know). It is the reason for the request, and it is determined so that the AI can calculate how to answer the request. Intents are commonly used with NLP systems.
Intentless (Starting to be in Use)	This is an Aisera term used to describe the fulfillment process where Intents are not required because the system uses LLMs to intelligently review the Title and Description of a Workflow or Hyperflow and then determine the mandatory and optional input data fields needed to execute the flow. This process of conversationally collecting required inputs is called slot filling. See also: Using Workflow Descriptions.
Large Language Model (LLM)	The Large Language Models are built and trained by companies like Google, Meta, and Open AI that are able to store and train AI models using massive amounts of data to understand and generate human language. LLMs, such as GPT, Claude, or Llama, use a character prediction-algorithm to take a user-input (prompt) and respond by predicting the characters that should follow the provided prompt. These LLMs are typically leveraged for tasks like writing text, summarizing content, translating languages, or powering conversational agents. These models are so good with language, that you can tell then what persona to use when talking with customers and what type of conversations you want them to have. However, you still have to train the model to use your specific key words, domain language, and ontology to achieve the best outcome. So there are maintenance tasks at the beginning where you need to review unresolved Requests and optimize the bot for your specific use cases. In the context of Conversational AI or GenAI: LLMs serve as the core engine behind natural, flexible, and context-aware dialogue. Unlike traditional intent-based models (such as ICM), LLMs can generalize to unseen inputs, allowing more open-ended conversations without needing explicitly trained phrases.
LLM Gateway (not currently available in Admin UI)	Aisera’s tool for connecting to 3rd party LLMs such as Bedrock or Llama.
Natural Language Understanding (NLU)/Natural Language Processing (NLP) This is used in Aisera Conversational AI 1.0, but may be phased our of the Aisera Admin UI in future versions - if all customers upgrade to Conversational AI 2.0 and opt for Intentless conversations. (Still in Use)	In NLU and NLP models, the AI uses algorithms that understand the natural language in user Requests. This enables the models to undersrant, interpret, and generate human language. The original Gen AI applications, such as GPT-2 and GPT-3 used these models. Intents: What the user is trying to accomplish with their bot Request. See also: Configuring Intents and Fulfillment. Phrases: In the Aisera Gen AI Platform, Phrases are added to Intents to trigger the Intent. They represent different ways that a bot user can ask for the same result. For example, "get tickets", or "get ticket details" might be some of the phrases used to get ticket-related details. ICM: Intent Classification Model, the model we used in Conversational AI 1.0 to match a user’s utterance to their intent, based on the overlapping phrasing. Example: Utterance = “Hey can I get some help with a pwd reset?” Phrase = help with ${password} reset ${password} = entity w/ synonyms: Pwd Password Pass-word passkey Passphrase pw Intent = Identity Management : Password Reset
Next Best Actions	See Agent Assist and Next Best Actions.
Ontology	A set of concepts and categories in a subject area or domain that shows their properties and the relations between them. In the Aisera Gen AI platform, the Ontology feature is a structured representation of knowledge that includes a taxonomy of categories and their associated entities, along with their properties. Defines not just what things are but how they relate to each other—entities, synonyms, classes of entities. It supports supports entities, categories and multi level hierarchy of entity categories. Entity: A term that is relevant to the domain and customer use case. These include: acronyms, nouns, and noun phrases. These are words object, typical nouns, but can also include REGEX functions and other verbiage classifications Synonyms: Terms that are used interchangeably in similar language contexts. There are 3 types of synonyms in the Aisera Gen AI platform: Orthographic variants (two-step verification, 2SV). Morphological variants (biometric authenticators, biometrics authenticator). Semantic variants (e-signature, online signature, virtual signature). You can import ontology files or run a Generate Ontology Job to create the ontologies from Knowledge Base Articles. See also: Ontology.
Prompts Studio (not currently available for Customer Admins)	A tool in the Aisera Admin UI that allows you to manage LLM / Prompt usage throughout the solution you’re building. Prompt studio allows you to test the performance of various prompts / tasks against different models/versions. It includes the full version history / disaster recovery to minimize the risk from user modifications to prompts.
Prompts Library (not currently available in Admin UI)	The Prompts Library, within the Prompts Studio, includes optimized prompts for different tasks and domains. If you’re using Conversational AI 2.0, you can select prompts from this library to use in your application/bot.
Public KB (can’t be set as a Fulfillment source by Customer Admins yet)	A Public KB is a Knowledge Base that can be viewed by anyone using a browser on the internet. You can set your application/bot to use these public sites as a Fulfillment Source. Since the Aisera platform is associated with both Knowledge Bases and Knowledge Base Articles, Knowledge Base Articles are abbreviated in the Product Documentation as KBAs, and Knowledge Bases are abbreviated as KBs.
Retrieval-Augmented Generation (RAG)	A software indexing technique that uses an LLM to index external documents and add that indexed information (as data chunks) to augment the training that the LLM has already received. This leads to more accurate and relevant responses. In the Aisera Gen AI platform, you can generate a RAG Index Job after you have added and Integration and Data Source to your Aisera tenant instance and then run a Data Ingestion Job. Use the Settings > System Jobs menu to setup and run a RAG Indexing Job. When you run a RAG indexing System Job, you can choose the following options: Knowledge Graph - structured representations or knowledge that organize enities (nodes) and their relationships (edges) into a machine-actionable and human-readable way (workflow). See also: How to Create a Knowledge Graph. Vector Search - a technique that uses mathematical representation of data, called vector embeddings, to find related itesm KoGNER/KGNER - Knowledge Graph distilled for Named Entity Recognition. A framework that integrated knowledge graphs into Named Entity Recognition models to enhance their performance. Choose this option if you want to generate indexes from your Knowledge Base Articles. See also: System Jobs and How to Schedule a Knowledge Base Crawl.
Request	The input entered by a bot user after the initial greeting.
Requests Window	The Requests window gives you the ability to view the global data for the Requests that people have sent to your bot (or that you have sent to your bot as a test), as well as the ability to open specific Requests and look at the detailed metrics. See also: Requests Window.
Request Analyzer	The Aisera Gen AI platform includes a service that analyzes all conversations, including unresolved conversations, and applies auto-tags to all Requests so the models can sort them as related to Intents, Workflows, Knowledge Documents, or Internal Server/ Flow Execution Errors categories. This conversation service is known as the Request Analyzer. The resulting Conversation Status can be: Resolved Unresolved Casual Assisted Abandoned Unhandled Reviewing the status of your conversations gives you the ability to optimize the conversation components and improve the accuracy of your application/bot. See also: Request Analyzer.
System of Record (SOR)	The System of Record (SOR) is a designation that you can give to Tickets, Knowledge Base Articles, or Service Catalog items that allows the Aisera Gen AI to write information from the Aisera internal vector data store back to your original Data Source. The Aisera Service User for your external Data Source needs both Read and Writer permissions to the original data source tables related to the objects that you have designated as SOR items. If you don’t use use cases (like Agent Assist or Ticket Concierge), your Aisera Service User just needs Read permission for those tables. That Aisera Admin UI includes a menu item that gives you a view of the items in you SOR.
Taxonomy (Still in Use)	A tree-like classification system that organizes Ontology entities into classes / parent-child relationships. Taxonomy has been de-emphasized along with NLU / ICM. However, customers are still using it. Example: Vehicle (Synonyms: automobile, motor vehicle) Car (Synonyms: passenger car) Sedan (Synonyms: saloon, 4-door) SUV (Synonyms: sport utility vehicle, crossover, 4x4) Truck Pickup 18-Wheeler
Ticket	Any Case or Issue that is created using a customer’s Customer Relationship Management (CRM) ticketing system.
Ticket Concierge	Ticket Concierge is part of a group of Tickets Operations that are often used with the Agent Assist or Webchat bots. Ticket Concierge was created to resolve the customer issues proactively. When a user creates a Ticket, the Aisera Gen AI platform will listen to the Ticket Creation Event and send a resolution to the Tcket via the email channel/bot. Early versions of Ticket Concierge used the Listener channel, while later versions use the Event Studio. NOTE: Ticket Concierge sends an email to the Ticket that is added as a Ticket Comment. It does not send email to the bot user. The Ticket Concierge feature gives you the ability to define the following within a flow: Ability to define a separate pipeline for the channel which might need to be different from the app-level pipeline. Ability to provide different responses, if users clicked on Thumbs Down and mark Ticket Concierge answers not. helpful. Ticket Concierge can now add Conversation History directly as a comment to the ticket. This serves the purpose of delivering comprehensive context to agents. Now, agents can gain a detailed understanding of how users have interacted with the proposed solution provided by Ticket Concierge and thus streamlines the troubleshooting process. See also: Ticket Concierge with Event Studio.
TRAPS	Aisera provides a proprietary security system that uses a Trusted, Responsible, Auditable, Private, and Secure (TRAPS) framework to protect your applications against attacks. See also: Security & Privacy (TRAPS).