Platform · AI Agents

AI agents that work
inside your processes.

Configure agents with a model, instructions and the exact tools they're allowed to use. They run on triggers and schedules, call your functions and APIs, hold real conversations, and answer plain-English questions about your data — returning live tables and charts.

Grounded on your own entities, permissions and tools — never a generic chatbot.
2 tools granted
Structured output
app.processo.io / agents

Agents

Configure AI agents for your process

Search agents…
R
revenue_analystREFLEX

Answers revenue questions over invoice and payment entities.

O
order_flowORCHESTRATOR

Routes new orders, checks stock and raises approvals.

D
doc_readerREFLEX

Reads uploaded PDFs and fills entity fields automatically.

S
support_botREFLEX

Drafts replies on WhatsApp threads with order context.

revenue_analystREFLEX Cancel Save
General Tools Execution Memory
Basics
revenue_analyst REFLEX claude_sonnet
Turns plain-English revenue questions into SQL over the invoice, customer and payment entities of this process.
System Prompt
You are the revenue analyst for this process. Use the supplied entity metadata, column descriptions and example queries to answer the question with a single SQL statement. Cast JSON attributes before aggregating — eg sum((entity_data->>'amount')::numeric). Return only SQL, no explanation, and never modify data.
Agent configuration

Define exactly how each agent thinks and acts.

An agent is a saved configuration, not a black box. You pick the language model, write the system instructions that shape its behavior, and grant it a precise set of tools — your functions, queries and connectors. Nothing more. It only ever sees the entities and rows your permissions allow.

Model & instructions

Choose the model, set temperature, and write the persona, rules and guardrails the agent must follow.

Scoped tool access

Attach functions, saved queries and connectors. The agent can only call what you've explicitly granted.

Memory & permissions

Conversations carry context across turns, and every action respects your RBAC and row-level access.

Finance Copilot · Configuration
Model
claude_sonnet temp 0.2
Instructions
You are a finance assistant. Answer using only the
Invoices, Customers and Payments entities. Always show
amounts in USD and flag anything overdue > 30 days.
Tools & functions
query_invoices aging_report send_whatsapp
Runs on
Conversation On record create Daily 9:00 schedule
How it works

From a question to a real action — in one loop.

Agents follow an agentic loop: understand the request, decide which tool to call, run it against your data, then reason over the result to answer or act.

1

Trigger

A user message, a workflow event (create / edit / status change), or a scheduled job wakes the agent.

2

Plan

The model reads your instructions and decides which of its granted tools and functions to call.

3

Act

It executes the tool — a saved query, a function, or a connector like WhatsApp or Stripe — within permissions.

4

Respond

Results are reasoned over and returned as a clear answer, a table or a chart — or kick off the next step.

Natural-language queries

Ask your data anything. Get answers, not dashboards.

Agents translate plain English into queries over your entities, then render the result the way a human would want to see it. No SQL, no report builder, no waiting on analysts.

Grounded on your model. Questions resolve against your real entities, fields and relationships — answers are specific, not hallucinated.

Tables & charts on demand. "Break revenue down by region" returns a chart; "list overdue invoices" returns a table — automatically.

Follow-ups keep context. "Now just the ones over $20k" refines the previous answer without repeating yourself.

Respects access. Each user only ever queries the rows and fields their permissions and hierarchy allow.

app.processo.io/agents/revenue-analyst
Agent Assistant Revenue Analyst
Break revenue down by region for last quarter
Assistant10:42 AM 8.6s
Thinking
West leads at $412k — up 18% on the prior quarter:
Revenue by region · last quarter$ recognized
412k
318k
276k
203k
141k
WestEastCentralSouthIntl
Send this breakdown to the regional heads.
Question 1 of 1
Which channels should this breakdown go out on? *
Email
WhatsApp
SMS
Slack / Teams
Select all that apply
Back Submit answers
Ask something…
Agentic actions

They don't just answer — they get things done.

Because agents call your real functions and connectors, they can take action on the result, end to end.

Reach out

Send a WhatsApp template, an email, or a Slack message — fired straight from a workflow the agent triggers.

Update records

Create or edit records, change a status, or move an item through your process — every write checked against permissions.

Call external systems

Hit REST APIs, charge through Stripe, read from entities — using secrets from the secrets manager.

Run on a schedule

Wake every morning to chase overdue invoices, summarize yesterday's tickets, or post a daily pipeline digest to Slack.

React to events

Trigger on create, edit or status change — a new lead, a stuck ticket, a high-risk order — and act within seconds.

Converse two-way

Hold a real conversation in the inbox — text, images, audio, documents — keeping context across the whole thread.

In the wild

One platform, agents for every process.

Collections agent

Runs daily, finds invoices overdue more than 30 days, and sends each customer a WhatsApp reminder with the amount and due date.

Support triage agent

On every new ticket, classifies priority, suggests a category, and routes high-severity issues to the on-call queue.

Sales copilot

Answers "which deals slipped this month?" with a table, then drafts follow-up emails for the reps who own them.

Works with

Agents are wired into the rest of Processo.

Ship your first agent

Give your processes an AI teammate.

Pick a model, write the instructions, grant a few tools — and let it answer questions and run on your business while you sleep. No code, no infrastructure.

Build an agent See pricing