A channel for Digital Onboarding

A complete loan application, inside WhatsApp

A borrower who has to download an app, create an account and find a scanner will usually not finish. Bankify runs the whole loan application inside a conversation the customer already has open: browse products, get a real priced quote, enrol, photograph the documents, and accept the terms with a PIN. Nothing about it is a chatbot demo — every step writes to the same records a branch application would.

Request a demo Compare all four modules

Part of Bankify Digital Onboarding. Bankify sells four modules — this is one way of using one of them, not a separate product.

A real quote before you ask who they are

Most lending bots demand an ID number in the first message, and most customers leave. Bankify is deliberately identity-last: an anonymous person can list your products, name an amount and a term, and get a full priced quote before being asked for anything personal. Identity is requested at the point they decide to apply.

Browse available loan products without identifying yourself
Give an amount and a term, get a priced quote back in the chat
Per-band rates and fee amounts shown separately, alongside the estimated monthly payment
The amount actually financed shown where fees are capitalised into the loan
Identity is asked for at the apply step, not the greeting

The quoted figure is the figure that books

This is the part that matters and the part most conversational lending gets wrong. The bot does not estimate. It calls the same pricing service that builds the repayment schedule when the loan is booked, using the same calculation kernel, so the number quoted in chat is the number the customer signs for. The assistant is explicitly forbidden from computing a repayment itself.

Quotes come from the lending engine, not from the language model
The same calculation kernel prices the quote and builds the booked schedule
The assistant is prohibited from estimating a repayment in conversation
Pricing reflects the product’s real interest method and fee treatment

Enrolment without leaving the conversation

A new customer registers inside the chat. Where the channel supports a structured form the details are captured in one, otherwise the assistant collects them conversationally and parses what the customer types. Either way it produces the same customer record a branch would create.

Name, national ID number, date of birth, phone, gender and address
A structured WhatsApp form where one is published, conversational capture where it is not
National ID and proof of residence uploaded as photographs
Both documents can be deferred at enrolment and supplied later
Enrolment finishes by setting a PIN

Documents, photographed and checked on arrival

Customers photograph documents rather than scanning them, and the accepted evidence reflects what borrowers can actually produce. Each upload is read on arrival and checked against the profile — advisory checks that fill in a reviewer’s checklist, never an automatic decline.

Identity, proof of address, proof of income and passport photograph
Proof of address accepts a utility bill, bank statement, lease, employer letter, sworn affidavit, or a councillor’s or headman’s letter
JPG, PNG or PDF, up to 10 MB
Documents checked on upload for type, ID match, freshness and income consistency
A check that cannot complete becomes "needs manual review" rather than a rejection

A PIN as the signature, with its own freshness window

Acceptance is not a "yes" in a chat window. Confirming a loan requires a PIN entered fresh at that moment — a much shorter window than the ordinary session — so the act of signing is distinct from merely being logged in. Where the channel supports it, the PIN is entered in a masked form rather than typed into the conversation.

A PIN entered within a five-minute window is what accepts the loan
That window is separate from, and far shorter than, the ordinary session
Masked entry where the channel supports it, so a PIN never sits in the chat transcript
Three attempts, then a lockout
PIN setup, change and reset all self-service, with reset verified against date of birth

The full agreement, behind the same PIN

Alongside the summary in chat, the customer gets a link to the complete agreement on the web, protected by the same PIN. It is a consumer-protection disclosure rather than a marketing summary, showing what the credit actually costs.

Itemised fees and the amount actually financed
Level instalment, total interest and total cost of credit
An effective annual cost rate, not just the headline rate
Rendered under the lender’s own name and letterhead, not Bankify’s
Reachable only with the customer’s PIN

It becomes a real application, with a real reference

Submission is not a notification to a staff member. The conversation creates and submits an application into the same state machine a web application enters, and returns the reference the backend generated. If the backend does not return one, the assistant says the submission failed rather than inventing a reference.

Creates and submits into the same application state machine as every other channel
Returns the real backend reference, and refuses to fabricate one on failure
Enters the same approval, KYC-snapshot and audit path
Forwards on approval to Bankify Loan Management or your existing LMS

What else the conversation handles

The same assistant covers the routine servicing questions that otherwise reach a call centre, which is what makes the channel worth maintaining once the application is done.

Loan and application status, balances and the next payment due
View and update profile, address and employment details
View and add a bank account for disbursement
KYC status, including what is missing or expiring
Logging a complaint, routed into the complaints register
Corporate applications, where a business borrows and a verified representative acts for it

Not only WhatsApp

The same conversation engine runs on other channels, so the work of configuring products, requirements and templates is not spent on a single messenger.

WhatsApp, including structured forms and template messages
Facebook Messenger
An embeddable web chat widget for your own site, with the same anonymous-quote-then-identify flow
A mobile application API
On the web widget the PIN is handled out of band, so it never enters the message stream

Who it’s for

Microfinance institutions

Reach borrowers who will never visit a branch and will not install an app, using the messenger already on their phone.

Lenders with high form abandonment

Move the application into a conversation and let customers finish it across several sittings.

Lenders keeping their current LMS

Adopt the WhatsApp channel on its own; approved applications forward into the system you already run.

Where this fits in the platform

WhatsApp is one channel into Bankify Digital Onboarding; a web application form is another, and both produce the same reviewable application.
The quote given in chat comes from the same calculation kernel that books the schedule in Bankify loan management software.
Reaching borrowers without a branch visit is one of the main reasons institutions evaluate Bankify as microfinance software.
The same messaging channel carries automated repayment reminders once the loan is live, through loan collection software.

Frequently asked questions

Can a customer complete a whole loan application on WhatsApp?

Yes. They can browse products, get a priced quote, enrol as a new customer, upload their identity and supporting documents, and accept the terms with a PIN — all inside the conversation. It creates and submits a real application and returns the reference the backend generated.

Is the quote in the chat accurate, or an estimate?

It is the real figure. The assistant does not calculate anything itself — it calls the lending engine, which uses the same calculation kernel that builds the repayment schedule when the loan is booked. The number quoted in chat is the number the customer signs for.

Do customers have to identify themselves before getting a quote?

No. The flow is deliberately identity-last. Anyone can list your products and get a priced quote for an amount and term without giving any personal details. Identity is requested at the point they choose to apply.

How does a customer sign for a loan over WhatsApp?

With a PIN entered fresh at the moment of acceptance, within a five-minute window that is separate from and much shorter than the ordinary session. Where the channel supports it the PIN is entered in a masked form, so it never appears in the chat transcript. The full agreement is also available on the web behind the same PIN.

How is KYC handled when the customer is just sending photographs?

Each upload is read on arrival and checked for document type, whether the ID number matches the profile, whether the document is still fresh, and whether stated income is consistent with the document. Those checks are advisory — they populate a reviewer’s checklist. A check that cannot complete becomes "needs manual review", never an automatic decline.

Is WhatsApp Banking a separate Bankify product?

No. WhatsApp is a channel into Digital Onboarding, not a separate module. Bankify sells four modules, and an application that starts on WhatsApp lands in exactly the same place as one submitted through a web form.

What proof of address will the flow accept?

A utility bill, a bank or building-society statement, a lease, an employer letter, a sworn affidavit, or a councillor’s or headman’s letter. The list is deliberately broad, because a utility-bill-only rule excludes a large share of the borrowers our customers actually lend to.

Want to see this against your own loan book?

Talk to a product specialist about which modules fit your institution today.

Request a demo Contact sales