shimli
Book a demo
Operations

From the WhatsApp orderto the invoice issued.

Orders come in however the customer likes and go out invoiced without anyone keying them in by hand.

Distribution and wholesale · high-volume ordering

Three colleagues reviewing a process together in a meeting room
The starting point

Customers ordered however they could. The system accepted one format only.

How orders came in

WhatsAppphoto of a handwritten list
Emailspreadsheet with 60 line items
Phone calldictated and written down
WhatsAppa two-minute voice note
Someone keyed it into the systemItem by item, checking code and price against the catalog.
0orders came in every day
0people transcribing full time
0 hbetween the order and its invoice

The bottleneck wasn’t selling or shipping: it was turning four ways of ordering into one way of recording. And every transcription was a chance to get a code, a quantity or a price wrong, which later showed up on the invoice.

How it runs

From reading the order to the issued invoice.

  1. 01Order intake

    It arrives by WhatsApp, email or phone.

  2. 02Reading and structuring

    It extracts customer, product and quantity.

  3. 03Validation against the catalog

    Code, price and available credit.

  4. 04Entry into the system

    The order enters the ERP already structured.

  5. 05Invoice and notice

    It’s issued and the customer gets a copy.

What it uses from Shimli

The pieces that make this case work.

AI agents

They read the order however it arrives and return it structured.

  • How it was used here
  • They read the photo, the spreadsheet, the dictation and the voice note alike.
  • If a product isn’t in the catalog, they ask before entering it.
  • They confirm the order with the customer before it enters the system.
Orders · DistributionAgent · actionslive
leerPedidoReadExtract product, quantity and customer
validarCodigoDatabaseCheck every code against the catalog
aplicarPrecioDataApply the price agreed with that customer
revisarCreditoLogicVerify the available credit line
cargarERPActionWrite the structured order
Code InterpreterWeb Search

Flow Builder

The order’s journey, from intake to invoicing.

  • How it was used here
  • Trigger: a message arrives from a registered customer.
  • Branches by amount: above a threshold it goes through approval.
  • If the customer exceeds their credit line, the flow stops and raises a flag.
Flow canvas
Start
BranchContact · 5068…Contact · 5066…Else
AI AgentCrediAmigoDigital assistant · live

Databases

Catalog, prices and customers each line is validated against.

  • How it was used here
  • The catalog validates the code, not the memory of whoever types it.
  • Each order keeps its origin: channel, time and the original message.
  • Customer-specific pricing applies on its own.
Databaseoptica · crediamigo Edit structure
NamePhoneIDAmountStatus
Cliente 01+504 ····0801-····$75,000active
Cliente 02+504 ····0501-····$75,000active
Cliente 03+504 ····1623-····$75,000review
Cliente 04+504 ····0107-····$75,000active

Integrations & API

The order enters the ERP and the invoice goes out the same way.

  • How it was used here
  • It writes the order into the ERP with the fields that system expects.
  • The invoice is issued in the usual system, not a new one.
  • The customer receives their copy through the same channel they ordered on.
CRM
Core
Bureau
WhatsApp
FusionAI
AI Voice
SMS
Email
API
POST /v1/agents/run200 OK
The outcome

What changes in the operation.

01The order comes in however it arrives

The customer keeps ordering by WhatsApp, email or phone. The difference is nobody transcribes it anymore.

02The catalog validates, not memory

Code, price and credit are checked against the database before the order exists, not later on the invoice.

03Invoicing stops waiting

The invoice no longer depends on someone finishing yesterday’s typing queue.

This case describes how the agent operates on the process. Result figures are defined with you during discovery, against your own indicators.

The other three processes

The agent runs the same way on every one of them.

CollectionsFrom cold receivables to confirmed payment.See the case CreditFrom the incomplete file to a decision with everything in view.See the case SupportFrom the message queue to the reply that resolves.See the case
Common objections

What teams usually ask before getting started.

Straight answers to the most common questions from credit, collections and technology teams.

Shimli isn't a chatbot. It's a platform to build AI agents that understand context, analyze information, execute tasks in your systems and escalate to a human when needed. A chatbot replies; an agent resolves.
The agent never decides out of bounds. It operates under policies, rules and guardrails defined by your institution, and the final decision can still be human. You define what it can and cannot do.
No. You can start with a controlled pilot and connect progressively via APIs. Shimli doesn't require replacing your current systems, it integrates on top of what you already have.
The next step

Let's build this same case
on your operation.

We review your operations process, identify where the agent fits and define with you the indicators we're going to move. No commitment, nothing to install.

Design my operations case

From $139 a month, with a 14-day trial and self-serve setup. See pricing