# D2L Brightspace + Agentic OS Integration > Source: https://ibl.ai/resources/integrations/d2l-brightspace-agentic-os *Deploy purpose-built AI agents directly inside D2L Brightspace — on your infrastructure, under your control.* ibl.ai Agentic OS connects with D2L Brightspace to give institutions a full AI agent platform that runs on their own infrastructure. Build, deploy, and manage specialized AI agents without surrendering data ownership or accepting vendor lock-in. Unlike generic AI add-ons, Agentic OS agents are purpose-built with defined roles — advising, tutoring, grading support, content delivery — and wired directly into Brightspace workflows via LTI and open APIs. Every agent your institution deploys stays on your servers, governed by your policies, and compliant with FERPA, HIPAA, and SOC 2 requirements from day one. ## Capabilities ### Purpose-Built Agent Deployment Design and deploy AI agents with specific roles — advisor, tutor, grader, onboarding guide — directly inside D2L Brightspace courses and org units. ### Real-Time Learner Event Processing Agents respond instantly to Brightspace activity events such as quiz submissions, content views, and discussion posts to deliver timely, contextual interventions. ### Automated Grade Passback Agent-assessed activities and AI-scored interactions write results directly back to the Brightspace gradebook via LTI Advantage grade services. ### Institution-Owned Infrastructure All agent logic, conversation data, and model configurations run on your servers. No learner data leaves your environment without your explicit authorization. ### Multi-Agent Orchestration Coordinate multiple specialized agents across departments and courses from a single Agentic OS control plane, with shared context and unified audit logs. ### Open API Extensibility Extend agent capabilities by connecting Agentic OS to Banner, PeopleSoft, SIS platforms, or custom institutional data sources alongside Brightspace. ## Setup ### Step 1: Provision Agentic OS on Your Infrastructure (~2–4 hours) Deploy the ibl.ai Agentic OS runtime on your cloud or on-premises environment. Use the provided Helm charts or Docker Compose templates to complete the base installation. Requirements: - Kubernetes cluster or Docker-compatible host - Minimum 8 vCPU, 32 GB RAM for production - Outbound HTTPS access to D2L Brightspace tenant ### Step 2: Register an OAuth 2.0 Application in Brightspace (~30 minutes) Create a new API application in the D2L Brightspace admin panel to generate client credentials. Assign the required scopes for users, enrollments, grades, and content. Requirements: - D2L Brightspace System Administrator access - OAuth 2.0 client ID and secret ### Step 3: Configure the Brightspace API Connector (~30 minutes) Enter your Brightspace tenant URL and OAuth credentials into the Agentic OS admin console. Run the connection test to verify read and write access to course and user data. Requirements: - Brightspace tenant URL - OAuth 2.0 client credentials from Step 2 ### Step 4: Register Agentic OS as an LTI 1.3 Tool (~1 hour) Add ibl.ai Agentic OS as an external learning tool in Brightspace using LTI Advantage. Configure deep linking and grade passback to enable full two-way integration. Requirements: - LTI 1.3 tool registration credentials from Agentic OS - Brightspace LTI tool admin permissions ### Step 5: Build and Configure Your AI Agents (~2–8 hours depending on agent complexity) Use the Agentic OS visual builder to define agent roles, knowledge bases, and trigger conditions. Assign agents to specific Brightspace courses, org units, or user roles. Requirements: - Agentic OS admin access - Defined agent use cases and knowledge sources ### Step 6: Test, Monitor, and Go Live (~1–2 days) Run agents in sandbox mode within a Brightspace test course. Review logs in the Agentic OS dashboard, validate data flows, then promote agents to production org units. Requirements: - Brightspace sandbox or test org unit - Agentic OS monitoring dashboard access ## Technical Requirements **D2L Brightspace Requirements:** - D2L Brightspace version 20.22.6 or later - System Administrator access for API and LTI configuration - OAuth 2.0 API application registration enabled - LTI Advantage (1.3) tool registration support - Brightspace Data Sets (BDS) access for batch analytics **Agentic OS Infrastructure Requirements:** - Kubernetes 1.24+ or Docker Engine 20.10+ on customer-managed infrastructure - Minimum 8 vCPU and 32 GB RAM for production deployment - PostgreSQL 14+ for agent state and session persistence - Outbound HTTPS (port 443) to Brightspace tenant URL - TLS 1.2 or higher for all inter-service communication **Compliance and Security Requirements:** - Data residency must align with institutional FERPA obligations - Role-based access control (RBAC) configured in Agentic OS admin console - Audit logging enabled for all agent interactions and data access events - SOC 2 Type II controls documented and enforced at infrastructure level ## FAQ **Q: Does the D2L Brightspace and Agentic OS integration require data to leave our institution's servers?** No. Agentic OS is deployed on your own infrastructure. All agent logic, learner interaction data, and model configurations remain within your environment. Only standard API calls are made to your Brightspace tenant. **Q: Which version of D2L Brightspace is required for this integration?** The integration requires D2L Brightspace version 20.22.6 or later with LTI Advantage (1.3) and OAuth 2.0 API application support enabled. Most current Brightspace tenants meet these requirements. **Q: How is Agentic OS different from D2L's built-in AI features?** Agentic OS lets institutions build and own custom AI agents with defined roles and institutional knowledge. Unlike built-in AI tools, agents run on your infrastructure with no vendor lock-in and can connect to external systems like Banner or PeopleSoft. **Q: Is this integration FERPA compliant?** Yes. Because Agentic OS runs on your infrastructure and all data flows remain within your environment, the integration is designed to support FERPA compliance. ibl.ai also supports HIPAA and SOC 2 compliance requirements. **Q: Can we deploy multiple AI agents for different departments or courses in Brightspace?** Yes. Agentic OS supports multi-agent orchestration. You can deploy distinct agents for advising, tutoring, onboarding, and corporate training, each scoped to specific Brightspace org units, courses, or user roles. **Q: How long does it take to set up the D2L Brightspace and Agentic OS integration?** The core technical integration — infrastructure provisioning, API connector setup, and LTI registration — typically takes one to two days. Building and configuring individual agents varies based on complexity and use case. **Q: Can Agentic OS agents write grades back to the D2L Brightspace gradebook?** Yes. The integration uses LTI Advantage grade services to pass agent-assessed scores and activity completions directly back to the Brightspace gradebook in real time or on demand. **Q: Does ibl.ai support integration with other systems alongside D2L Brightspace?** Yes. Agentic OS is designed to connect with multiple systems simultaneously. You can integrate Brightspace alongside SIS platforms like Banner or PeopleSoft, HR systems, and custom institutional data sources through open APIs.