New — AI Agents

Your processes now have their own intelligence.

Autonomous agents that observe case events, analyze data and documents, and take real actions — change stage, create tasks, review documents, enforce deadlines. No prompt. No code. No integration.

AI Agents available from the Starter plan.

Are you still doing this manually?

Reading every submitted document to check if it's correct

Remembering to follow up on pending items before the deadline

Writing summaries every time a case changes stage

Manually setting up initial tasks for each new case

Your processes generate dozens of events per day. Documents submitted, stages changing, deadlines approaching. Each event requires an action. Today, that action depends on you remembering.

Marketplace

Meet your AI team

Each agent is a specialist with its own persona, memory, and reasoning. Install from the marketplace in one click.

Summarizer

Case Summaries

Meticulous analyst that never misses a detail.

Generates executive summaries automatically when the case changes stage or at scheduled intervals. Ideal for keeping stakeholders informed without manual effort.

Triggers

Stage changeScheduled interval

Ações

Create commentGenerate summarySchedule next execution

Doc Reviewer

Document Review

Detail-oriented reviewer that analyzes every page.

Analyzes every document submitted to the case, identifies completeness or compliance issues, and automatically creates correction tasks.

Triggers

Document submitted

Ações

Read documentsCreate commentCreate taskReview document

Deadline Guard

Deadline Control

Proactive monitor that never forgets a deadline.

Monitors SLAs and deadlines for each case. Sends alerts before deadlines expire, adjusts priorities, and escalates when needed.

Triggers

Scheduled intervalCase created

Ações

Read tasksSend notificationAdjust prioritySet deadline

How it works

01

Choose and install

Open the Agent Marketplace in template settings. Choose from 9 ready-made agents. Click "Install".

02

Configure triggers and permissions

Define which events activate the agent (document submitted, stage changed, deadline passed). Choose exactly what it can read and do.

03

The agent works on its own

On each event, the agent analyzes the case context, reasons about what to do, and executes real actions. Everything logged in the timeline.

No code. No integration. No prompt. The agent already knows what to do.

Not a chatbot. An agent that acts.

Chatbots / Copilots
CaseFy AI Agents
How it works
You ask
The agent observes
Context
What you type
Entire case: stage, fields, docs, people
Actions
Replies with text
Changes stage, creates task, reviews document
Memory
Forgets each session
Remembers between executions
Activation
Manual (prompt)
Automatic (22 triggers)
Setup
Prompt engineering
One click in the marketplace

30+ actions. 22 triggers. Full control.

What agents can READ

  • Case data (fields, stage, status, priority)
  • Documents (with full text content)
  • Complete process timeline
  • People and companies involved
  • Tasks and their statuses
  • Conversations and comments
  • Scheduled activities

What agents can DO

  • Add comments and reply to threads
  • Create and assign tasks with deadlines
  • Move the case to a different stage
  • Fill fields automatically
  • Send notifications and emails
  • Generate executive summaries
  • Review documents with feedback
  • Adjust case priority and deadline
  • Manage tags
  • Schedule next execution (wake-up)
  • Store memory between executions

You control exactly what each agent can read and do. Granular permissions per agent, per template.

22 events that activate your agents

Case

Case createdStatus changedCase assignedCreated via public form

Stage

Entered stageLeft stage

Field

Field value changed

Artifact

Document submittedForm submittedForm approvedForm rejectedTask completedReview completedDecision made

Time

SLA breachedDeadline reachedScheduled intervalInactivity

Person

Person addedPerson removed

Signature

Signature completedSignature rejected
Segurança

AI with control, not on autopilot

Granular permissions

Each agent receives only the capabilities you allow. Read and write configured separately.

Complete audit trail

Every agent action is logged in the case timeline. Author, date, context, and result.

Automatic circuit breaker

3 consecutive errors = agent automatically paused for 1 hour. No risk of infinite loops.

Configurable limits

Maximum actions per execution, maximum executions per day. You set the limits.

Modelos

Choose the AI model per agent

OpenAI

GPT-5 Nano
GPT-5 Mini
GPT-5.2

Anthropic

Claude Haiku 4.5
Claude Sonnet 4.5
Claude Opus 4.6

Available models vary by plan. Starter: GPT-5 Nano. Professional: all standard. Enterprise: all including premium.

AI Agents in every plan

Starter

R$79/usuário/mês

1 marketplace agent, 200 credits/month, GPT-5 Nano

Prove the value with one agent.

Mais Popular

Professional

R$159/usuário/mês

Unlimited agents, 2,000 credits/month, all standard models

Full intelligence for your operation.

Enterprise

Sob consulta

Everything unlimited + custom agents + premium models

For operations that demand the best.

View all plans

Extra credits available: Starter R$39/1,000 | Pro R$25/1,000 | Enterprise R$15/1,000

Your processes are waiting for intelligence.

Start free. Once on Starter, activate your first agent in 2 minutes. No code, no complex setup.

Up to 3 free users forever. No credit card required.