Growing processes, teams that need to work together, and AI Agents taking on more tasks all raise practical questions. What is connected? How does each action happen? Where should control remain? For people already using Pipefy, these questions are part of expanding everyday operations.
The October 8 Pipefy Updates event brought together product improvements around these needs. This recap covers the main developments for governing operations, orchestrating processes, and running processes with AI, with dedicated articles to help you explore what matters to your work.
For people governing operations
Helping teams move faster while monitoring costs, security, and stability is a connected responsibility. As a company expands its use of AI, it needs to choose models that suit different tasks and understand how its processes consume resources.
Pipefy AI Lite, Standard, and Pro expand model choice for individual agent behaviors. Bring Your Own Model, or BYOM, also lets companies connect their own models, including options from Anthropic, and use different models according to their needs.
Company Overview brings together usage indicators across multiple Pipefy organizations. Integration Usage Stats provides a view of integration consumption by pipe and flow. These insights help teams investigate where usage is concentrated before making decisions about capacity and resource allocation.
Visibility also extends to auditing and monitoring. The Pipe Audit Logs API provides access to audit records by pipe, while webhook alerts send information about automation and AI Agent executions to the tools used by the team. Version history lets teams save a stable pipe configuration and restore it when a change requires a return to the previous state.
Read more: For people governing operations: AI with control.
For people orchestrating processes
When process knowledge stays with the person who built it, taking over operations or assessing a change requires reconstructing that context. With AI Agents acting within the flow, understanding what happens becomes even more important.
The improvements let you follow the same pipe at three levels. Across the ecosystem, Map shows connections to other pipes and databases. Card Volume adds traffic information to those connections, helping you distinguish busy routes from connections with no volume during the period.
At the structural level, Flow View brings phases, automations, and AI Agents together in a diagram. At the card level, Timeline shows the history and identifies who performed each action. This sequence helps connect the process design to what happened with a particular request.
Pipefy Integrations completes this view with protected credentials, execution monitoring, and step testing in the editor. Bulk operations support creating or updating up to 10,000 cards in an import and exporting cards to a file using the required filters.
Read more: For people orchestrating processes: see what happens.
For people running processes with AI
Reading a lengthy contract, checking information in another system, and applying a calculation rule are different tasks, even when they belong to the same request. Execution needs to combine the appropriate capabilities and preserve people's involvement in sensitive decisions.
Behavior and automation testing lets teams validate scenarios before publishing. Improve instructions supports the review of agent guidance with suggestions that users can accept, reject, or edit.
During execution, MCP Client can retrieve information from systems such as Salesforce or HubSpot and connect actions to tools such as Slack. Calculations & Analysis executes business rules through code, while Max Effort supports deeper reasoning to analyze lengthy documents and populate fields according to the agent’s instructions.
Human Validation can route decisions, such as a reimbursement above a company defined limit, for approval by email. Logs help teams follow execution steps and the AI Credits consumed by each action. For technical teams, MCP Server also supports querying cards and accessing Pipefy tools through compatible assistants and development environments.
Read more: For people running processes with AI: test and track.
You can also browse the Pipefy product roadmap to find releases and supporting resources. Return to the articles whenever you need to revisit a topic at your own pace.
