Medellín, Colombia · UTC−5 · Remote operation across Latin America and the United States hello@quarl.co EN · ES
quarl ES

Chatbots for businesses of every size

It answers on WhatsApp or on your site, using your real prices and hours, books appointments and takes orders. And when it does not know something, it says so and hands the conversation to someone on your team.

2,000,000 active users served · 10 countries in operation

  • WhatsApp Business API
  • Website
  • Instagram
  • Books appointments
  • Takes orders
  • Voice notes
  • Hands off to a person
Contact us See the channels Free, no pitch · the proposal lands in 48 hours

Built on

  • Anthropic
  • Claude
  • Google Gemini
  • Google Cloud
  • LangChain
  • LangGraph
  • PostgreSQL
  • Qdrant
  • Datadog

An engineering team that has already been on the other side.

Ninety seconds: who we are, how we work and what you get at the end.

The problem

Whoever answers first gets the sale

Most messages do not arrive while you are sitting in front of your phone. They arrive on a Sunday, at nine at night, or while you are serving someone else. And whoever is asking about a price almost never writes to one business: they write to three, and buy from whoever answers first.

That is the real cost, and it shows up in no report. It is not the unanswered message: it is the sale that already closed somewhere else before you opened the conversation.

During the day the opposite happens and costs the same. Eight in ten messages are the same dozen questions, usually hours, price, location and whether it is in stock, and someone on your team is typing them out by hand, one at a time, instead of serving the customers who actually need a person.

Channels

Where the assistant lives

01 3–5 weeks

WhatsApp

The channel where customers actually write in Latin America. Built on the official WhatsApp Business API, with voice note transcription and escalation to your team.

02 3–4 weeks

Website

Widget or full page, aware of what the user is looking at, with handover to a human agent without losing the thread.

03 5–8 weeks

Inside your product

An assistant embedded in your application, with the user’s session, permissions and data.

04 4–6 weeks

Internal for your team

Over documentation, procedures and internal systems. Usually the highest-return one, and the least often requested.

Always included

Five things it comes with, at no extra charge

  • It speaks with your real data.

    Your prices, your hours, your stock, your return policy. Not generic text somebody wrote once and left to go stale.

  • Live lookups when needed.

    "Has my order shipped?", "do you have it in medium?", "what do I owe?". It connects to your system and reads the current value, not last week’s copy.

  • It hands off without making the customer repeat.

    When someone from your team steps in, they see the whole conversation from the first message. The customer never has to explain twice.

  • Every month it tells you what it could not answer.

    That list is the most useful thing you will get: it shows what is missing from your information and what people are asking without finding an answer.

  • It gets measured, so you know if it improves or degrades.

    We build a set of real questions with their correct answers and re-run it every time something changes. Without that, nobody knows whether the assistant is better or worse than yesterday.

In practice

What it does, depending on what you sell

01

Restaurant or delivery

“What’s on today?” · “how late are you open?” · “do you deliver to my area?”

What the assistant does

Shows today’s menu, confirms whether it covers the area and takes the order with address and payment method. Anything off-script goes to a person.

02

Store or online shop

“Do you have this one?” · “in medium?” · “has my order shipped?”

What the assistant does

Checks stock and shipment status live. If it is sold out, it offers what is actually in the warehouse instead of letting the conversation die.

03

Clinic, salon or workshop

“Any opening tomorrow?” · “how much is it?” · “where are you?”

What the assistant does

Shows open slots, books, sends the reminder and rebooks whoever cancels, which is the money that normally just disappears.

04

Real estate or vehicle sales

“Is it still available?” · “how much down?” · “do you take trade-ins?”

What the assistant does

Filters by budget and area, sends photos and the listing, and hands the agent only the leads who already answered what mattered.

05

Field services or construction

“How much would it cost?” · “when can you come?”

What the assistant does

Asks what it needs in order to quote (what, where, how much and by when) and hands the team a complete case instead of a bare “hi, info”.

This is where the data comes in

An AI system is worth what its sources are worth. These are the standard connectors; anything with an API or a database connects the same way, and what has no API is handled by file.

SAP

Enterprise ERP

ERP

Oracle

ERP and database

ERP

NetSuite

Cloud ERP

ERP

Salesforce

CRM and service

CRM

HubSpot

CRM and marketing

CRM

PostgreSQL

Database and pgvector

Databases

Fourteen sectors, and in each one the measurement changes

Insurance

Policy wordings, exclusions, claims

Says what a policy covers, quoting the clause and the version in force that day. If the wording changed in March, it answers with March’s.

AI for Insurance →

An example

— Am I covered if my car is stolen from a public car park?

Yes, with a 10% deductible. 2026 policy wording, clause 4.3.

A RAG system in production: three years, two million users

We built and operated the assistant for a loyalty platform serving more than two million active users across ten countries.

We built the full pipeline: document ingestion and normalization, chunking, embedding generation, vector store on Azure AI Search and Pinecone, and retrieval with grounded generation on LangChain.

We held it above 99% availability for three years.

RAG systems →

Frequently asked questions

01

Is my business too small for this?

It comes down to one thing: whether your information changes. If it is fewer than twenty FAQs and the answer is almost always the same, do not hire us: a monthly subscription platform solves that for a fraction of what a build costs, and that is the recommendation that comes out of the first call. Where it does make sense is a large catalog with many variants, prices or stock that move, or a case where a wrong answer costs you a customer or a legal problem. That is where generic platforms start confusing similar products and inventing, and that is engineering. If you are not sure which case you are in, the fifteen-minute call is enough to find out.

02

How is this different from using ChatGPT?

ChatGPT knows everything except your company: not your prices, your inventory, your warranty policies or an order’s status. Asked about those it answers generically or invents. What we build searches your information first, answers only with what it found, and cites the source so you can verify it.

03

Can we use our existing WhatsApp number?

Yes, an existing number can be migrated to the WhatsApp Business API, though the process means it stops working in the regular WhatsApp app. That is why many companies prefer dedicating a new number to the assistant and keeping the current one for personal contact. It gets decided on the first call based on how your team works.

04

What does Meta charge for messages?

Since July 2025 Meta charges per delivered template message, not per conversation. Replying within 24 hours of the customer’s last message is free; what you pay for are the messages the business starts —marketing, utility, authentication— at a per-country rate. An assistant that only answers whoever writes in generates almost no Meta cost; one that follows up or sends campaigns does. You pay that directly to Meta, not to us, and we estimate it against your real volume before you sign.

05

Why the official API and not a cheaper automation?

Because unofficial methods get your number banned by Meta, and losing the number your customers write to is a hard disaster to reverse. The official API charges for the messages the business starts, but it is the only way to build on WhatsApp without risking the channel.

06

Does it understand voice notes?

Yes. They are transcribed through a speech-to-text service and processed like a written message. It is especially relevant in Latin American markets and is usually badly handled by generic platforms. There is a small per-minute transcription cost included in the monthly estimate.

07

What if the customer realizes it is a bot?

They will, and that is fine: the assistant introduces itself as an assistant from the first message. Trying to pass it off as human goes badly when discovered. What people dislike is not talking to a bot, but talking to one that solves nothing and will not let them reach a person. That is why escalation is a core function, not an extra.

08

Does it work in multiple languages?

Yes. Current models handle Spanish and English well, and the system can reply in the language it is written to. What needs care is that the knowledge base is in the right language: if your documentation is only in Spanish, English answers will be translations and should be validated before enabling them.

09

What does it cost to run per month?

Model consumption plus the vector store, roughly USD 150 to 900 monthly for a mid-market company, plus Meta’s template-message cost if it is WhatsApp. We do not charge a subscription on the system: it is yours, with its code and documentation.

Fifteen minutes. A concrete answer.

It gets fixed, it gets built, or it is not worth it. And if the diagnostic does not reach three actionable findings, it is not charged.

Message on WhatsApp