Bizagi vs Camunda (2026)

Agents act through AI Actions. A builder links connectors already set up in the project, SAP BTP connectors or MCP servers reached over server sent events, picks which of their actions the agent may use and writes the goal into the prompt, and the agent decides when to call each one, with up to 100 actions in total. MCP servers authenticate with an API key sent as Basic, Bearer or a custom header. Agents can gather more data, act in outside systems and start processes in Bizagi or third party systems this way.

Sourcehelp.bizagi.com/platform/en/ai_actions.htmread 2026-10-13

Agents act through the BPMN activities a builder places in an ad hoc sub process, so any service task, connector call or sub process becomes a tool the model can choose. Camunda offers more than 400 prebuilt connectors, including Salesforce, SAP OData and RFC, HubSpot, Slack, Microsoft Teams, GitHub, Twilio, WhatsApp, Google Drive and Sheets, Amazon S3, DynamoDB and Lambda, plus generic REST, SOAP, GraphQL and SQL. Credentials sit in secrets, a connector SDK builds custom connectors, and MCP client connectors let an agent discover and call tools on outside MCP servers.

Sourcedocs.camunda.io/docs/components/connectors/out-of-the-box-connectors/available-connectors-overviewread 2026-10-12

Agents run inside BPMN processes built in Bizagi Studio, as an action on an AI task that fires when a case enters or leaves it, from a button on a form, from entity forms or inside data pipelines, and Master and Persona entities can now run an agent as an Entity Action with its own input and output mappings. Several agents can run in one process, side by side or one after another, and their outputs map to typed fields in the data model, so later steps, business rules and people use them like any other case data.

Sourcehelp.bizagi.com/platform/en/ai_agents.htmread 2026-10-13

Processes run on the Zeebe engine as BPMN, with gateways, timers, events, sub processes and DMN decisions. An AI agent is an addressable loop inside that process. The model picks the next activity from an ad hoc sub process, the engine runs it as a governed step, and the result goes back to the model until the goal is met. Several agents can share one process, agents built in LangGraph, Amazon Bedrock or custom code can be connected and tracked as agent instances, and predictable steps run straight through while agents handle the judgment calls.

Sourcedocs.camunda.io/docs/components/agentic-orchestration/ai-agentsread 2026-10-12

An agent on an AI task runs whenever a case reaches that task, with no one asking. Cases themselves start from a timer start event on a schedule, from a message or signal thrown by another process, from events Bizagi consumes off Kafka topics, or from a person in the work portal, and AI Jobs in the Management Console schedule and monitor AI work on their own. Agents can in turn trigger events that start processes in Bizagi or in other systems, and processes can publish events back to Kafka.

Sourcehelp.bizagi.com/platform/en/start_events.htmread 2026-10-13

Processes start or move on without anyone pressing a button. Inbound connectors listen on an HTTP webhook, Kafka, RabbitMQ, Amazon SQS and SNS, Amazon EventBridge, Slack, email, Microsoft 365 mail, GitHub and Twilio, and BPMN timer and message events start or resume instances on a schedule or when a message arrives. An MCP start event registers a deployed process as a tool that outside AI clients can call, and each call starts a new instance.

Sourcedocs.camunda.io/docs/components/connectors/out-of-the-box-connectors/available-connectors-overviewread 2026-10-12

Enterprise Knowledge holds catalogs that are processed, indexed and served through retrieval augmented generation while agents run, kept in a vector database that serves retrieval and never trains a model.

File Knowledge builds a catalog from content extracted from documents and serves AI Agents, AI Workers and the Ask Ada assistant, and Bizagi Data Knowledge builds one from records stored in Bizagi entities for AI Agents. Catalogs sit in categories in Studio's AI Hub and are reindexed after a deployment or update.

Agents can also find conceptually similar past cases without exact keyword matches, and the knowledge can extend or replace an agent's default instructions with a company's guidelines and procedures.

Sourcehelp.bizagi.com/platform/en/enterprise-knowledge.htmread 2026-10-13

Agents can draw on company knowledge kept in a vector store. The Vector Database connector splits documents into chunks, embeds them with Amazon Bedrock Titan, Azure OpenAI, Google Vertex AI or OpenAI models, and writes them to Amazon OpenSearch, OpenSearch, Elasticsearch, Azure AI Search or Azure Cosmos DB. At run time it retrieves the most relevant passages by semantic similarity, so an agent can look up policies and procedures, search a large product catalog, reuse answers to past support questions or quote the exact policy behind a decision. The store grows as new documents are embedded.

Sourcedocs.camunda.io/docs/components/agentic-orchestration/rag-ai-agentsread 2026-10-12

Each agent call works from inputs mapped from the case and the knowledge catalogs attached to it, and writes its outputs back to case fields, so the case carries the state while it runs. Inputs can run to 500,000 characters, so whole documents and long case histories can be passed in on each call. An agent keeps no memory of its own that it reads from one case to the next.

Sourcehelp.bizagi.com/platform/en/ai_agents_faqs.htmread 2026-10-13

Each agent keeps a short term conversation inside its process instance, passing the last 20 messages to the model by default while always keeping the system prompt. Those messages can live in a process variable, in Camunda document storage with a time to live set to the process's expected lifetime, in Amazon Bedrock AgentCore Memory, or in a custom store on self managed setups. Recall from one process instance to the next relies on an outside service such as AgentCore Memory, which AWS runs.

Sourcedocs.camunda.io/docs/components/connectors/out-of-the-box-connectors/agentic-ai-aiagent-subprocessread 2026-10-12

AI Workers take on tasks people would normally do in one of three modes. In supervised processing the worker does as much of the task as it can and hands it to a person to review or finish, conditional processing follows rules the customer sets to decide when it works alone, and autonomous processing falls back to a person whenever it cannot finish. Each decision comes with its reasoning for a person to check, and agent outputs land in form fields that people review before a case moves on.

Sourcebizagi.com/en/platform/ai-workersread 2026-10-13

People sit in the process as user tasks, and Tasklist shows each assigned task whether the step before it ran automatically or an AI agent produced the result a person now acts on. A user task can be one of an agent's tools, so the agent asks for help or sign off inside its loop. For MCP tools, a gateway in front of the client lets listing and read operations through and routes selected calls to a user task for confirmation before they run. Task work can also reach people in Microsoft Teams, and teams can build their own task apps on the API.

Sourcedocs.camunda.io/docs/components/connectors/out-of-the-box-connectors/agentic-ai-mcp-client-human-in-the-loopread 2026-10-12

Bizagi holds ISO/IEC 27001:2022, its cloud for US government is FedRAMP Moderate, and independent reviews found the cloud meets HIPAA and HITECH requirements. Users sign in through SAML, OAuth or LDAP with two factor authentication, role based segregation of duties and delegated administration govern who can do what, and roles can be kept in step with Microsoft Entra ID. Data is encrypted in transit and at rest, each customer's applications and data are isolated, and an annual transparency report covers government requests for customer data.

Sourcebizagi.com/en/platform/complianceread 2026-10-13

Camunda runs two identity layers, the Orchestration Cluster admin for runtime work and the management identity for Hub and Optimize, and both connect to an outside OpenID Connect provider such as Entra ID for single sign on, with LDAP and Active Directory on the licensed editions. Authorizations, roles, groups, mapping rules and tenants decide who can start, see and change processes, tasks and other resources. The company holds SOC 2 Type II, ISO 27001 and TISAX and is listed with the Cloud Security Alliance on its trust center.

Sourcetrust.camunda.comread 2026-10-12

Tracing in the Management Console records AI agent invocations, the requests and responses of the connectors an agent calls, tagged with the case number, and the responses AI Workers produce, at error, warning or information level, and AI Workers show the rationale behind each decision.

Traces are kept for 90 days and are meant for troubleshooting, with the platform's audit and accountability controls kept apart from them. An AI consumption report traces usage to processes, tasks, entities and applications and shows the models and tokens behind each execution, and an AI Metrics dashboard tracks agent use over time by process and by agent.

Sourcehelp.bizagi.com/platform/en/tools-tracing.htmread 2026-10-13

Operate shows each agent instance's state, such as thinking or calling a tool, its usage metrics and its conversation history grouped by loop iteration, including the tools it picked and the results it got back, for native and connected external agents alike. Health indicators and limit warnings flag an agent that keeps calling the same tool or nears its configured limits. Every process and decision instance can be inspected step by step, audit log APIs record user and user task actions, and Optimize builds reports, dashboards and KPI alerts.

Sourcedocs.camunda.io/docs/components/agentic-orchestration/evaluate-agents/monitor-ai-agentsread 2026-10-12

Each customer gets a dedicated cloud setup with isolated storage and its own URL, provisioned in the Azure geography it picks at onboarding, US, EU, Asia or US Government, and every service and resource stays inside that geography. Bizagi chooses the exact Azure region within it. Production data is replicated to a paired region and backed up nightly, with a 5 minute recovery point and a 3 hour recovery time.

Sourcehelp.bizagi.com/platform/en/available-regions-for-bizagi-c.htmread 2026-10-13

SaaS customers choose the region for each cluster. On Google Cloud the choices include Belgium, London, Iowa, South Carolina, Montréal, Toronto, Singapore and Sydney, and on AWS they include Frankfurt, Paris, Ohio, Oregon, Virginia and Singapore, each paired with a secondary backup region. Enterprise customers can ask for custom regions. Self managed Camunda runs the whole stack on the customer's own infrastructure, and agent model calls can route through a private endpoint inside the customer's network.

Sourcedocs.camunda.io/docs/components/saas/regionsread 2026-10-12

AI agent templates deploy straight from Bizagi Studio, with 13 for banking and investments, 7 for customer service and 10 for manufacturing and retail, grouped into predefined Solutions a builder can search by Solution, title or description. They cover jobs such as generating and comparing earnings reports, categorizing customer incidents, obfuscating personal data, composing incident responses, writing knowledge base articles, translating policies, comparing vendors and summarizing RFx responses. Bizagi Copilot helps refine a template's prompt in plain language.

Sourcebizagi.com/en/platform/ai-agent-templatesread 2026-10-13

Camunda Marketplace carries model blueprints and connectors from Camunda, partners and the community, installed straight into a Hub project or downloaded for the desktop modeler. AI agent blueprints such as AI Agent Chat Quick Start and AI Agent Chat With Tools run out of the box on the Camunda provided LLM in SaaS. Partner and community connectors sit outside Camunda's commercial support, and administrators of a self managed Hub can switch the Marketplace off.

Sourcedocs.camunda.io/docs/components/hub/workspace/modeler/modeling/camunda-marketplaceread 2026-10-12

Agents run on Azure OpenAI through Bizagi's private GPT service. For each agent a builder picks a model from those OpenAI currently offers, which change with OpenAI's updates and deprecations, choosing between direct response models and reasoning models with a set reasoning level, and tunes temperature, top p, frequency penalty and presence penalty. Files as inputs need GPT 4o or a newer model. Every choice stays within OpenAI's models, with no option to bring another maker's model or a customer's own deployment.

Sourcehelp.bizagi.com/platform/en/ai_agents_from_scratch.htmread 2026-10-13

The builder picks the provider and the backend for each agent. Anthropic Claude models run directly, through AWS Bedrock or through Microsoft Foundry, OpenAI models run directly, through Foundry or through any OpenAI compatible endpoint, and Google Gemini runs directly or through Google's Enterprise Agent Platform, with custom endpoints for private network setups and a custom provider on self managed installs. SaaS adds a Camunda provided LLM with a free budget for testing, so teams can switch between models such as Claude Sonnet 4.6 before moving to their own account.

Sourcedocs.camunda.io/docs/components/connectors/out-of-the-box-connectors/agentic-ai-aiagent-model-providersread 2026-10-12

The Bizagi API is a RESTful service on OData 4.0 with OAuth 2 bearer tokens, so outside applications can start cases for a process, search cases and work with process data. A JSON file in Swagger 2.0 and OpenAPI 3.0 format describes every OData endpoint for middleware, a full service index lists them, and examples include C# code. Since the Summer 2026 release a REST API also exports and imports deployments, so a team can automate releases from its own pipeline, and API traces can be switched on for debugging.

Sourcehelp.bizagi.com/platform/en/api_odata.htmread 2026-10-13

The Orchestration Cluster REST API drives processes, instances, user tasks, decisions, identities and audit logs, with SDKs for Java, TypeScript, Python, C# and PHP and the c8ctl command line tool. The Orchestration Cluster MCP Server opens operational work such as incidents, user tasks and process instances to AI clients under the same authentication and authorization model as the REST API. Older Operate and Tasklist REST APIs remain available for existing integrations.

Sourcedocs.camunda.io/docs/apis-tools/orchestration-cluster-api-mcp/orchestration-cluster-api-mcp-overviewread 2026-10-12

Builders try an agent in a Test Prompt window, which runs one input file at a time, shows which actions the agent called, and can differ from what happens at run time. AI Workers in supervised mode learn from each user after the 50th interaction on a task, building a personal policy from that user's behavior and feedback, and a team can switch that reinforcement learning on or off per task. Bizagi has retired its Autotesting tool, and there is no saved test set, quality gate or measure of how the learning changes results before an agent goes live.

Sourcereleasenotes.bizagi.com/en/release-notes/releases/fall-2025read 2026-10-13

Camunda Process Test runs an agent process as an integration test, keeping the real model in the loop while mocking outside tool calls. Conditional behavior reacts to whichever tools the agent activates instead of one fixed order, and judge assertions send an output and a plain language expectation to a model that rates the match against a set threshold, alongside semantic similarity checks. Tests run through a Spring Boot starter inside the project's own build.

Sourcedocs.camunda.io/docs/components/agentic-orchestration/evaluate-agents/test-ai-agentsread 2026-10-12

Screen level work runs through RPA bots from partners such as Automation Anywhere, which handle manual tasks while Bizagi manages the flow of data between people and robots. The bots and their runtime come from the RPA vendor, and Bizagi's own agents do not drive a browser or desktop themselves.

Sourcebizagi.com/en/solutions/robotic-process-automation/automation-anywhere-and-bizagiread 2026-10-13

Camunda RPA runs Robot Framework scripts on a standalone RPA worker to operate legacy mainframe applications, desktop applications such as ERP systems and websites that expose no API. Each script runs as a step in a process, so screen work on an old system sits in the same flow as API calls, agent steps and user tasks, and scripts are built and tested before they move to production workers.

Sourcedocs.camunda.io/docs/components/rpa/overviewread 2026-10-12