A customer says "hi" on WhatsApp and gets a five-insurer comparison back in the same chat. We built the rate database, the quotation engine and the Bitrix24 integration behind it.
The challenge
Every enquiry started the same way. A customer messaged asking for a quote, someone from the team replied to gather the details, then went away to look up rates across a dozen insurers held in a dozen separate Excel sheets, worked out the premiums by hand, and typed a comparison back. It took hours, sometimes a day. By then the customer had often gone somewhere faster. The sheets were also a maintenance problem on their own: when an insurer revised rates, whoever had the newest file quoted correctly and everyone else quoted from stale numbers.
Short version: a customer messages the broker on WhatsApp. A Gupshup chatbot asks what it needs to know. A webhook we built turns those answers into a comparison of five insurers, sends it back in the same chat, and writes the whole thing to a Bitrix24 deal. Quotes that used to take hours now take the length of a conversation.
What the team was doing before
Someone messages asking for a quote. A person replies, asks the questions, waits for answers.
Then the real work starts. Open a dozen Excel sheets, one per insurer. Find the right category and sum insured band in each. Work out the premiums. Type a comparison back into WhatsApp.
Hours, usually. Sometimes the next day. By then the customer had often quoted somewhere else.
The sheets caused their own problems. When an insurer revised rates, whoever had the newest file quoted correctly. Everyone else quoted from stale numbers and nobody knew until a customer queried it.
Where the enquiry starts now
The customer sends “hi”. That is the entire entry point.
The Gupshup chatbot takes over and asks whatever the broker has set up in their Gupshup dashboard. Asset details, coverage, sum insured, and so on. The broker controls those questions themselves. Changing what gets asked does not involve us.
When the conversation ends, Gupshup fires a webhook at our service with the answers.
The chatbot was the easy half
A set of answers does not produce a quotation. You still need to know what each insurer charges for that exact combination. That knowledge was sitting in spreadsheets.
So the first thing we built was not the bot. It was a database.
We took the insurers’ rate sheets and turned them into a structured store: companies, product categories, sum insured bands, and the parameters each insurer prices on.
Then we built a custom admin dashboard on top of it, and this is the part the client actually cares about. Their team adds an insurer, edits a category, updates a sum insured band, changes what a company covers. No code. No ticket to us.
When rates change, one person updates the dashboard. Every quotation after that uses the new numbers. One source of truth instead of a dozen files with different edit dates.
How does the quotation get built?
Our webhook takes the WhatsApp answers and queries the rate database. It finds which insurers cover that profile, calculates the premium for each against their own parameters, and ranks them.
The customer gets back the five best-matched companies, in WhatsApp, in the same thread they started.
Five was a deliberate choice. One quote gives the customer nothing to judge against and reads like a sales pitch. Twenty is a spreadsheet nobody opens on a phone. Five shows the market honestly and is still short enough to decide from.
And it all lands in Bitrix24
While the customer is reading the comparison, the same webhook is writing to Bitrix24. A deal gets created with the customer’s details and their answers, and the generated quotation is attached to it.
Nothing gets rebuilt by hand. The sales team opens the deal and sees what was asked, what was quoted, and when. Follow-up automation runs off that deal like any other.
What changed
Quotes go out at conversation speed, at any hour, whether or not anyone is at a desk.
Every comparison comes from the same rate source, so two customers with the same requirement get the same answer.
And because the broker maintains the database themselves, the system does not rot the moment insurers revise their rates. That was the thing we were most worried about at the start, and it is the reason we built the dashboard instead of hard-coding the rates.
Frequently asked questions
Can a WhatsApp chatbot create insurance quotations automatically?
Yes. The chatbot collects the customer's requirements, and a webhook calculates premiums against a rate database and sends the comparison back in the same chat. The chatbot gathers; the pricing logic sits behind it.
How do you connect Gupshup to Bitrix24?
Gupshup sends the completed conversation to a webhook, which processes the answers and writes to Bitrix24 through its REST API — creating the deal, populating fields, and attaching the quotation.
How are insurer rates kept up to date?
Through a custom admin dashboard where the broker's own team manages insurers, categories, sum insured bands and product parameters. No developer involvement when rates change.
Why compare five insurers instead of showing all of them?
Five is enough for a genuine comparison and short enough to read on a phone. Longer lists get skimmed or ignored.