> ## Documentation Index > Fetch the complete documentation index at: https://braintrust.dev/docs/llms.txt > Use this file to discover all available pages before exploring further. # Glossary This glossary defines key terms and concepts used in our product and documentation. ## Alert An automation that notifies you when a specific condition occurs on your logs in Braintrust. [Automations guide](/admin/automations) ## Automation A configured workflow that lets you trigger actions based on specific events in Braintrust. For example, sending an alert, or batch exporting data. [Automations guide](/admin/automations) ## Benchmark An evaluation designed to assess model performance across specific capabilities or against industry standards. ## Brainstore The high-performance data engine backing logs, search, and tables. [Brainstore blog post](https://www.braintrust.dev/blog/brainstore) ## Classification The topic label assigned to a trace by a facet, indicating which cluster the trace belongs to. Classifications appear as columns in the logs table and can be filtered or queried with SQL. [How Topics work](/observe/topics#how-it-works) ## Configuration Project-level settings that define behavior for evals, experiments, and integrations. [Project configuration guide](/admin/projects) ## Dataset A versioned collection of pairs of inputs and (optional) expected outputs. [Datasets guide](/annotate/datasets) ## Evaluation / Eval An eval consists of a task, dataset, and one or more scorers or classifiers. Evaluations can be: * **Offline**: run a task on a static dataset with scoring functions. * **Online**: real-time scoring on production or test requests. [Evals guide](/evaluate/run-evaluations) ## Experiment An instance of an offline eval run. Scores a specific task run on a given dataset. [Experiment guide](/evaluate/run-evaluations) ## Facet A dimension of analysis applied to traces by the Topics pipeline (for example, Task, Sentiment, or Issues). Each facet uses a preprocessor and an LLM prompt to extract a short summary per trace. [How Topics work](/observe/topics#how-it-works) ## Human review An option to route evaluations or tasks to human reviewers instead of, or in addition to, automated scorers. [Human review guide](/annotate/human-review) ## Inline parameters Configurable options defined directly in your `Eval()` call using Zod, Pydantic, or plain dicts. Inline parameters become UI controls in the playground when running remote evals or sandboxes. To change the values, update them in your eval code. ## Log An instance of a live production or test interaction. Logs can include inputs, outputs, expected values, metadata, errors, scores, and tags. Scorers can also be applied to live logs to conduct online evaluations. [Logs guide](/observe/view-logs) ## Loop agent Braintrust's AI agent that can help you with evaluation-related tasks, like optimizing prompts and generating dataset rows. [Loop guide](/loop) ## Metric A quantitative measure of model performance (for example, accuracy, latency, or cost) tracked over time and across experiments. ## Model An AI system (typically an LLM) that can be evaluated or monitored with Braintrust. Models can be first-party, third-party, or open-source. ## Organization Your company or team "home" in Braintrust. It holds all your projects, members, and settings. [Organizations reference](/admin/organizations) ## OTEL OpenTelemetry: the instrumentation standard Braintrust uses to collect and export trace and span data from integrations. [OpenTelemetry guide](/integrations/sdk-integrations/opentelemetry) ## Playground An interactive space where you can prototype, iterate on, and compare multiple prompts and models against a dataset in real time. A playground can be saved as an experiment. [Playgrounds guide](/evaluate/playgrounds) ## Preprocessor A function that formats a trace into readable text before a facet's LLM extracts a summary. The default Thread preprocessor renders traces as conversation narratives. Custom preprocessors can adapt non-conversational trace shapes. [How Topics work](/observe/topics#how-it-works) ## Project A container for related experiments, datasets, and logs. Use projects to segment work by feature, environment (dev/prod), or team. [Projects guide](/admin/projects) ## Prompt The instruction given to an AI model. Prompts are editable objects you can version and reuse across experiments and playgrounds. [Prompts guide](/deploy/prompts) ## Prompt engineering The practice of designing, optimizing, and refining prompts to improve AI model outputs and performance. ## Regression testing Evaluations that ensure new model or prompt configurations maintain or improve upon previous performance benchmarks. ## Remote eval An evaluation that is executed on external or third-party systems or services, allowing you to evaluate tasks in environments outside Braintrust. [Remote evals guide](/evaluate/remote-evals) ## Saved parameters Centrally-managed configuration for evaluations. Create versioned parameter definitions in Braintrust and load them across evaluations for reusability, version control, and environment management. [Parameters guide](/evaluate/write-parameters) ## Scorer The component responsible for judging the quality of AI outputs. Scorers may be: * Rule-based code * LLM-based prompts as judges * Human reviewers [Scorers guide](/evaluate/write-scorers) ## Setting An organization-level preference or control, including user management, billing, and global integrations. [Organizations reference](/admin/organizations) ## Span A single segment within a trace, representing one operation (for example, a model call or tool execution) with its timing and metadata. [Traces guide](/observe/view-logs#examine-a-trace) ## SQL SQL queries for filtering and analyzing eval results, logs, and metrics. Braintrust also supports BTQL, an alternative pipe-delimited syntax. [SQL reference](/reference/sql) ## Structured output A defined format (for example, JSON or XML) that models must follow, enabling consistent parsing and scoring of responses. [Structured output guide](/evaluate/write-prompts#create-prompts) ## Task A single unit of work, typically composed of an input, output, expected result, and evaluation. Tasks often appear within dataset or eval detail screens. ## Thread view A visualization mode for traces that displays the interaction as a conversation thread, showing messages, tool calls, and scores in chronological order. Thread view is particularly useful for debugging LLM conversations and multi-turn interactions. [Traces guide](/observe/examine-traces#view-as-a-conversation) ## Timeline view A visualization mode for traces that displays spans as horizontal bars, where bar width represents the duration of each operation. Timeline view is useful for identifying performance bottlenecks and understanding execution flow. [Traces guide](/observe/examine-traces#view-as-a-timeline) ## Topic A cluster of similar facet summaries, generated by the Topics pipeline to label recurring patterns across your logs. Each trace is classified with the closest topic. Examples: "Refund requests," "API errors." [How Topics work](/observe/topics#how-it-works) ## Trace An individual recorded session detailing each step of an interaction: model calls, tool invocations, and intermediate outputs. Traces aid debugging and root-cause analysis. [Traces guide](/observe/view-logs) ## User feedback End-user inputs and ratings collected from production that inform model performance tracking and future evals. [User feedback guide](/instrument/user-feedback) ## Workflow A type of task that can be used in playgrounds. Consists of a chained sequence of prompts that automate complex workflows, where one LLM call's output feeds into the next. [Workflows guide](/evaluate/playgrounds#workflows)