Give Agent Research a question or a task, build application features in Workspaces, and train predictive models from your data. Connect reusable agents and published predictions to the applications you create in Onyx.
Agent Research
Investigate your data and create dashboards, scripts, and other resources with the permissions you grant.
Agent Builder workflow trees
Replace prompt chains with inspectable workflow trees where each step performs a deliberate action with tool access.
Model Builder
Map training data from saved queries, train neural networks, and publish models for application predictions.
Application Workspaces
Build with an AI coding agent, review changes, preview your app, and publish versioned releases.
Investigate data, create resources, and carry out permitted operations. Follow the progress and keep the findings, files, and context for your next task.
Select an available Onyx model or a connected personal Codex account, then choose a supported thinking level. Follow-ups and saved progress help you continue an investigation.
Start with read-only analysis, grant specific actions, or use Full Control for every action your current role permits. The dashboard remembers your selections for this database.
Allow the agent to create visualizations, dashboards, documents, scripts, or draft agents. Schema changes, data updates, and maintenance each have their own permissions.
Review concise answers and downloadable evidence. With the relevant grants, run saved scripts or set up recurring work. Manage ongoing schedules in the Scripts scheduler.
One platform, from data to application
Develop your application in Workspaces, put Agent Research to work, and bring your database and AI models into the same development process.
Build the application around your idea.
Start a web or mobile app, choose a sample with its own database, or import a GitHub repository. Give an agent coding tasks, review its changes, and try the application in a private preview.
Investigate, understand, and act.
Give an agent a question or a task. It investigates your data, runs analysis, and can create dashboards, scripts, documents, and other resources with the permissions you grant.
Give your application a connected foundation.
Model your data, define relationships, search records, and access your backend through APIs and SDKs. The database sits alongside the tools you use to build your application.
Turn database queries into predictions.
Build neural networks from saved Onyx queries. Map features and targets, configure the architecture, monitor training, and publish a model your applications can call.
Ready to integrate a published model? Read the Prediction API reference →
Upload documents and crawl domains to provide supporting context. Structured data remains the foundation for grounded, current answers.
Upload PDFs, spreadsheets, HTML exports, and knowledge articles so your agents can cite and enrich answers without turning documents into your source of truth.
Crawl a domain on a schedule and keep policy hubs, help desks, and marketing sites indexed as they change—without manual exports or sync work.
Index files and crawled pages so agents can retrieve relevant passages and return grounded responses backed by searchable references.
Blend documents with direct database querying so responses stay current and reflect the real state of your system—without ETL or stale snapshots.
Onyx agents don’t just retrieve text—they execute plans: query tables, traverse relationships, and produce reusable outputs.
Design agents as explicit workflow trees. Each node performs a deliberate step and can call tools to query, traverse, and compute over your Onyx data.
Agents can search tables directly—filter, aggregate, and compute on live structured data—then enrich with documents only when it adds value.
Traverse resolver-defined relationships to pull connected context across your model without stitching graphs over the network or relying on brittle joins.
Give agents tool access to your data with the same approvals, visibility, and audit trails you use across Onyx Cloud.
Produce reusable artifacts—not just answers: computed metrics, structured datasets, narrative explanations, and visualization-ready outputs.
Build agents that compute answers from live structured data, with documents used only as supporting context.
From support and analytics to investigations and briefings—agents compute on live data and produce reusable artifacts.
Answer operational questions by querying live structured data, then enrich with policy docs, runbooks, and knowledge articles when relevant.
Compute account health, adoption milestones, and next-step recommendations from live data—no manual dashboard building required.
Aggregate feedback, tickets, and usage analytics, then generate structured summaries and prioritized recommendations as reusable outputs.
Trace incidents through connected data, traverse relationships, and produce a clear narrative backed by the exact queries and references used.
Continuously ingest documents and crawl sites so agents can cite the latest materials—without turning documents into your primary data model.
Generate weekly KPI briefs that pair narrative insight with computed metrics pulled directly from live structured data.
Onyx makes structured data queryable by tools, relationships traversable by design, and workflows reusable—so agents can do real work.
Most AI platforms stop at document search. Onyx agents compute answers by querying live structured data and traversing relationships, then use documents as enrichment.
Define behavior explicitly with workflow trees. Each step is inspectable, reusable, and grounded in tool execution against your real data model.
Resolvers execute inside the database system. Graph attachment happens at the model level—without stitching graphs over the network.
Invoke agents through REST APIs, WebSockets, SDKs, background jobs, or application hooks—so AI becomes part of your system, not a separate tool.
Start an application in Workspaces, investigate your data with Agent Research, and connect agents and model predictions through Onyx APIs.