ATMs, conversational AI and collections

Talking to the bank's customers, from the cash machine to the overdue-loan call

Take a shopkeeper in Nagpur who banks with a large private bank. In one month, he deals with the bank in several places. He deposits cash at the branch. He takes cash out of an ATM. He asks the bank's app why a payment failed. And when he falls behind on his business loan, his phone rings with a reminder call. The voice on the call belongs to software, not to a person.

A bank talks to its customers in more places than almost any other business:

the branch
the cash machine, or ATM
the app
the call centre
the WhatsApp message
the reminder call, when a loan is overdue

This Market Segment covers the software built for those conversations in banking in particular. The same kinds of conversations, held by software built for every industry, are covered elsewhere. (More on them in Industry Vertical 15, Customer service and CX.)

Every company named here has posted software engineering jobs in India. Famous companies that don't actively hire software engineers in India are left out.

This Market Segment has two sub-segments:

The branch and the cash machine: the bank's physical channels

The conversation and the collection: the AI agents that answer the bank's customers and call its borrowers

The shopkeeper meets his bank at the branch and the ATM, its physical channels, and in the app, the call centre, WhatsApp and the reminder call about an overdue loan. Each place is labelled with the companies on this page that build software for it, and the numbers 1 and 2 match the two sub-segments below.

Keep reading, free

Two more sections are on this page: The branch and the cash machine and The conversation and the collection. Sign in to read them here, in full.

Continue with Google

The branch and the cash machine
The conversation and the collection

The branch and the cash machine

ATMs and self-service banking, branch and channel software

Take the shopkeeper's visit to the ATM from the start of this Market Segment. The ATM, and the software that runs it, come from a small number of companies. Banks increasingly pay one of them to run the whole fleet of ATMs for them, as a service.

NCR Atleos

with engineers in Hyderabad and Mumbai, has the most job postings in this Market Segment. It makes cash machines and runs ATM networks for banks. It sells the machine and its software, and, more and more, the whole fleet run as a service. NCR Atleos was separated from NCR in 2023. Its Hyderabad centre builds much of the software.

Impacto Digifin Technologies

with engineers in Bengaluru, sells banking software for branches and channels. A channel is any way a customer reaches the bank, such as the branch, the app or the phone.

Diebold Nixdorf, the other large ATM maker, is covered in Market Segment 19.1, Core banking and banking as a service (BaaS).

The conversation and the collection

AI assistants and voice agents for banks and credit unions, secure messaging, conversational collections, debt recovery software

Back to the shopkeeper's question in the bank's app: why did the payment fail? More and more, an AI agent answers it. An AI agent is software that holds the conversation and takes action by itself. A voice agent is one that talks on the phone.

interface.ai

with engineers in Bengaluru, sells an AI agent to American banks and credit unions. A credit union is a bank owned by its own members. The agent answers customers in chat and on the phone. It can check a balance, freeze a card and make a transfer.

Eltropy

sells credit unions secure messaging, video banking, and a contact centre with AI agents.

GreyLabs AI

with engineers in Bengaluru, sells speech analytics and voice agents to Indian banks and lenders. Speech analytics means software that listens to calls and analyses them. GreyLabs AI's software listens to calls in Indian languages.

Back to the shopkeeper's overdue loan. The rest of the companies here collect. Collections means getting overdue loans repaid. When a loan is overdue, the bank calls the borrower. Digital lending has made tens of millions of small loans in India. So in India, the reminder call is often made by software.

Skit.ai

with engineers in Bengaluru, sells voice agents that make collection calls.

Fundamento

with engineers in Delhi NCR, sells voice agents for collections and for onboarding, the steps of signing up a new customer, in Indian languages.

Resollect

with engineers in Bengaluru, sells software that manages non-performing loans through recovery and legal action. A non-performing loan is one on which the borrower has stopped paying for a long time, usually ninety days or more.

DPDzero

runs early collections for lenders as a service. Early collections means reminding a borrower in the first days after a payment is missed.

The collection call made by an agent.

The collection call is one of the largest uses of voice agents in India. It is frequent, it follows a script, and it is expensive when people make it. It is also where the rules matter most. The regulator limits when and how borrowers may be called, and the voice agent must follow those rules on every call.

When a loan payment is missed, early collections begin with reminder calls made by voice agents, and a loan left unpaid for a long time, usually ninety days or more, becomes non-performing and moves to recovery and legal action. Throughout, the regulator limits when and how borrowers may be called.

That is how a bank talks to its customers. The shopkeeper in Nagpur meets the bank at the branch and the ATM. He gets his answer from an AI agent in the app. And he hears his reminder from a voice agent that follows the regulator's rules.

Who they hire

Who these companies hire, and for what, is on What banking and regtech hires for.

← Previous in Banking and regtech

Fraud prevention, AML and regtech

Next in Banking and regtech →

none, this is the last