Deskferry
Build AI agents that handle recurring tasks across 1,500+ apps using plain English, no code, and built-in approvals.

DeskFerry is an AI automation platform for teams that want recurring work handled inside the apps they already use. It lets users describe a task in plain English and build an AI agent that can read, write, and take actions across more than 1,500 apps. It is designed to handle both quick everyday requests and longer, multi-step workflows.
The platform is aimed at the people responsible for getting work done, rather than requiring engineering involvement for every automation. Support, operations, sales, and founding teams can create agents that handle repeatable tasks, respond to events, and run on a schedule.
Key Features
Assist Mode
Handle everyday requests that only take a few minutes.
Assist mode is designed for smaller tasks where you need an AI helper to complete work quickly without building a larger automated workflow.
Agent Builder
Create agents for more involved, multi-step processes.
The Agent Builder can coordinate work across several applications, making it suitable for workflows that require multiple actions or systems.
Recurring Schedules and Event Triggers
Run work automatically when it needs to happen.
Agents can be scheduled to run repeatedly or triggered by events, allowing teams to automate processes that would otherwise require manual follow-up.
Human Approvals
Keep people involved before important actions happen.
Agents can pause for approval before sending messages, making updates, or spending money. This provides a control point for workflows where fully autonomous execution would be inappropriate.
Custom Skills
Give agents additional instructions for how work should be handled.
Custom skills can define things such as:
- Brand voice
- Output formats
- Artifacts
- Filing requirements
- Workflow-specific instructions
This helps agents produce work that follows the team's existing processes.
Knowledge and Memory
Give agents access to relevant information and ongoing context.
DeskFerry can use indexed documents and URLs as knowledge sources, while memory allows agents to retain preferences and context across runs.
Activity Logs
Review what happened during each agent run.
The activity history records executions and shows where an agent paused, giving teams visibility into automated work and approval points.
Multiple Access Channels
Run DeskFerry agents through different interfaces.
Agents can operate through:
- Web
- API
- Slack
This allows teams to trigger or interact with workflows through the tools they already use.
Built for AI-Powered Workflow Automation
DeskFerry is a good fit for:
- Founders
- Support teams
- Operations teams
- Sales teams
- Customer success teams
- Business teams
- Small companies
- Teams managing repetitive workflows
Common Use Cases
- Lead qualification
- Support ticket triage
- Invoice follow-ups
- Automated reporting
- Customer communication
- Sales operations
- Recurring administrative work
- Multi-step business workflows
- Document-based tasks
- Scheduled data processing
- Event-driven automations
- Cross-app workflow management
Why It Matters
Many business processes involve repetitive work spread across several applications. Teams may need to read information from one system, update another, send a message, create a record, and then follow up later. Traditionally, setting up this kind of workflow can require separate automation tools or engineering support.
DeskFerry combines these tasks around AI agents that can work across more than 1,500 applications. Users can describe what they need in natural language and build workflows without having to manage the underlying integrations themselves.
The platform also focuses on controlled automation rather than simply letting agents operate without oversight. Human approval steps, activity logs, knowledge sources, and memory give teams ways to control and review how agents operate.
It supports integrations with tools including Gmail, Slack, Notion, Salesforce, HubSpot, Zendesk, Jira, Stripe, and Google Sheets. Users can also choose models from providers including OpenAI, Anthropic, Google, and xAI without managing individual API keys inside DeskFerry.
For teams starting with automation, the combination of Assist mode and Agent Builder provides a path from small tasks to more complex workflows without needing separate systems for each level of automation.
Automate Recurring Work With DeskFerry
Describe a task in plain English, connect it to the apps your team already uses, and let AI handle recurring work through assistants and agents. DeskFerry supports scheduled and event-driven workflows, human approvals, shared knowledge, and activity tracking across more than 1,500 applications.