Great when the job is getting to a working flow quickly.
Visual agent builders are strong for composition, prompting, and iteration. They are not always the place teams want to manage production review and operational accountability.
Mermaduckle is the operating layer for AI workflows that touch customers, revenue, or internal systems. Build visual flows, route high-impact actions through human approval, manage agents and secrets, keep an audit trail for every run, and deploy the same Rust service in your own environment or with rollout support.
This domain includes a public sandbox at /app. Customer production
deployments can run as separate self-hosted or supported environments.
Most teams start with a tool that is optimized either for visual flow design or for broad automation across external services. The gap appears later, when an AI workflow needs review gates, auditability, runtime ownership, and a deployment model the infrastructure team can live with.
Visual agent builders are strong for composition, prompting, and iteration. They are not always the place teams want to manage production review and operational accountability.
Automation platforms win on connector breadth and event plumbing. They are a less natural fit when the core requirement is governed AI workflow execution.
The product centers on workflow runtime, approval, audit, agent operations, and a deployment model that can move from evaluation into customer-owned environments.
Public competitor positioning points to three dominant patterns in the market: connector-heavy automation, visual agent composition, and developer-first event orchestration. Mermaduckle is strongest in the operating layer those products usually leave to custom process.
Broad automation, connector coverage, and code-plus-visual workflow assembly.
Mermaduckle is the tighter fit when approval gates, audit history, and AI workflow ownership are the first-order requirement.
Rapid composition of agents, chains, prompts, RAG flows, and experiments.
Mermaduckle is the better operational fit when teams need review, reporting, runtime visibility, and a product built around governed execution.
API-first event automation, integrations, and fast developer workflows.
Mermaduckle is more focused when the problem is running AI workflows safely rather than maximizing integration and event breadth.
Workflow studio, approval queue, audit trail, agent operations, secrets, integrations, and deployment in one product surface.
Use it when the workflow itself is part of the product, the process, or the risk surface.
The live product is not a shell around one builder screen. It already includes the workflow, agent, approval, audit, settings, and deployment-adjacent surfaces teams use once workflows stay in production.
Create, edit, import, export, debug, and run workflows from one visual canvas.
Generate a draft workflow from a prompt, then refine it as an operator.
Route high-impact steps through explicit human review before continuing the run.
Inspect historical events, export audit output, and keep a durable run record.
Manage agent configs, test prompts, store secrets, issue API keys, and review health.
Trigger from external systems, run on cadence, and connect the workflow to the rest of the stack.
One of the current points of confusion was treating the public app on this domain and the self-hosted product as if they were different things. They are not. This site hosts a sandbox. The same service can be deployed separately for a customer-owned or supported environment.
Open /app to inspect the current workflow, approval, agent, and audit
surfaces without having to deploy your own environment first.
The repository includes the Fly deploy path, Dockerfile, and runtime configuration you need to launch a separate customer-owned instance.
Teams that want faster adoption can use a supported rollout for the first workflow, the first production environment, and the first operating handoff.
/ company and product overview/app live sandbox of the current product/docs deployment guide for self-hosted rollout/api/health service readiness endpointflyctl deploy -a <app-name> --remote-only --config fly.toml
The repo includes a multi-stage Dockerfile, GitHub Actions workflow, and PowerShell helper for customer-owned Fly deployments.
Mermaduckle is easiest to adopt when the buying model matches the product model: start in a self-hosted environment, then add help where production ownership becomes expensive. The point is not feature gating. The point is faster rollout, better operating discipline, and clearer support once workflows matter.
Use the open-source product for evaluation, internal pilots, and customer-owned deployments that your team wants to run directly.
Bring Mermaduckle into production with implementation help, operating guidance, and a faster path through the first governed workflow.
Once the workflow matters to the business, the relationship shifts toward response commitments, environment reviews, upgrades, and operational support.