Motor Olie — One step of the process, from the inside. The rule makes the work move.

Every process at Olie is a sequence of steps, and each step runs on an engine. The request comes in with a deadline and a goal, is only distributed with what it needs, is done by people and AI agents, talks to your systems and comes out measured, with the cost of each run.

  • 4 layers: rules, team, connections and measurement
  • 10 parts in each step
  • Not even the owner overrides the rule
The engine drawing

From the request to the next step. Everything that happens in between has a rule.

The orange card is the request: a refund request, a new hire, a proposal. It goes through two gates, one on entry and one on exit, and in between the step works with people, AI, deadlines and goals.

RequestSLA + KPIPRE-REQUISITECONNECTIONSAPIMCPERPWhatsAppSTAGEMOTOR OLIEAIAIPeople + AI agentsDeadline + triggerGoals + real costR$ 7,85GOVERNANCEnextOVERFLOWsame card, in parallelLogistics funnelFinance funnel
One step in Motor Olie. Illustrative values.
In short

What is Motor Olie? It is how Olie makes each step move by the rules.

Each step of a process in Olie Flow AI runs like this: the request comes in with a deadline and a goal, is worked on by people, AI agents or integrations and only moves forward once what was agreed is done. No exceptions, not even for the owner.

On entryDeadline, goal and complete data

The request arrives with an SLA and a KPI. The prerequisite gate only releases distribution with complete data.

In the middlePeople, AI and your systems

People and AI agents at the same step, with training, deadline and trigger. The step talks to API, MCP, ERP and WhatsApp.

On exitGovernance, overflow and cost

The card only moves forward with what was agreed filled in, overflows to other funnels in parallel, and goals, time and cost of each run are recorded.

The engine's logic

Four layers, ten parts. A refund request going through each one.

Example: a customer asks for her money back for a shoe whose sole came off. See what each part does with this request, without anyone having to chase anyone.

Layer 1

Rules

What is allowed, what isn't and when
  1. Piece 1 The rule chasesThe request arrives with a deadline and a goal

    The card is born with the SLA "start within 15 min" and the KPI "time to first response". Started in 38 s: within the SLA.

  2. Piece 2 The rule locksPrerequisites before distributing

    Order number, product photo and Pix key. The key is missing: nothing is distributed, and the AI asks the customer for the key on WhatsApp.

  3. Piece 3 The rule locksGovernance to move forward

    Without an assessment and a reason recorded, the card doesn't leave the step. Only the Customer Service team moves this card, and not even the owner overrides it.

  4. Piece 4 The rule chasesTime ran out? The trigger acts

    Target of 2 hours, limit of 8. Past the limit: it alerts the coordinator, the AI tells the customer the new deadline and the card becomes a priority.

Layer 2

Team

Who does it, knowing how to do it
  1. Piece 5 The rule teachesTraining inside the step

    Júlia gets the card through the step's rule and finds the video "How to review a refund", the checklist and an AI assistant for questions. Whoever joined yesterday executes the right way today.

  2. Piece 6 The rule coordinatesPeople and AI agents on the same team

    One agent checks the photo for a manufacturing defect, another cross-checks the customer's history. The AI prepares, Júlia approves. Hybrid approval, all in the history.

Layer 3

Connections

The step talks to the rest of the company
  1. Piece 7 The rule distributesExecutes or overflows

    Need to pick up the product and send the Pix? The same card starts running at the same time in the Logistics and Finance funnels, each with its own steps.

  2. Piece 8 The rule connectsConnects with your systems

    Approved: the bank sends the Pix, the ERP posts the reversal, the customer gets the receipt on WhatsApp and the board's assistant looks up the case through MCP.

Layer 4

Measurement

What happened, how long it took and how much it cost
  1. Piece 9 The rule showsGoals at each step

    Time target and limit, expected cost, volume (40 reviews per day) and quality (customer score ≥ 9). Each goal can trigger an action.

  2. Piece 10 The rule recordsCost and history of each run

    Who did it, when, how long it took and how much it cost, person and AI separately. In the example: Júlia's 11 minutes and a few cents of AI.

You design the step once. The engine makes the rule happen on every card, every day.

Try the engine

Choose the request and go part by part. Lock it, blow the deadline, run the agents.

Interactive simulation with three examples: sales proposal, customer refund and employee hiring. The tour runs on its own and stops when you touch something.

Loading the simulation…

You saw the engine run. Now bring your process.

30-minute meeting with the team, a ready-made template from the Marketplace or a chat with sales on WhatsApp right now.

Updated on · Source: Olie Flow AI technical knowledge base

Sales