---
title: "Takenos Case Study: AI Fintech Support | StudioChat"
description: "How Takenos automates 70-75% of fintech support with a StudioChat assistant that reads receipts, checks status systems and never exposes sensitive data."
url: https://studiochat.io/case-studies/takenos
locale: en
translations: { es: https://studiochat.io/es/case-studies/takenos.md, pt: https://studiochat.io/pt/case-studies/takenos.md }
site: https://studiochat.io
---

Case Study · Fintech · Support & Compliance

[Takenos logo](https://www.takenos.com)

LatAm

# A blurry receipt, a KYC question, a blocked transaction. All of it arrives through the same channel, resolved without touching sensitive data.

70-75%

Automation and deflection

20+

Scenarios covered

100%

Sensitive data never exposed to the model

3

New agents in expansion

"We reduced the handling time of every ticket to just one minute, without losing the judgment the team already resolved it with."

T

Takenos CX Team

Customer Experience, Takenos

The Challenge

## Claims, provider outages, and KYC — all under the same policy

Fintech support can't afford improvisation: every claim needs evidence, every provider outage generates tickets with no context, and every identity check has to follow policy to the letter.

#### Claims that need evidence

Photos, PDFs, and receipts arrived separately, and building a claim by hand meant back-and-forth with the user.

#### Provider outages

When a provider went down, every failed deposit or blocked transaction ended up as a manual ticket with no prior context.

#### Onboarding that couldn't improvise

Identity and account verification needed to follow policy to the letter, no matter who answered.

The Solution

## Joaquina: an agent that interprets, checks, and protects

StudioChat built Joaquina, an assistant integrated into the Intercom Help Center that interprets evidence, checks the right status systems, and never exposes sensitive data to the model.

#### Integrated into the Intercom Help Center

Answers inside the same Help Center the team already used, as another agent.

#### Interprets photos, PDFs, and receipts

Builds a complete claim from what the user already sent, without asking them to retype it.

#### Connected to the right status systems

Checks transactions, KYC, card limits, and accounts live — without exposing sensitive data to the model.

#### Creates work units, not just replies

When a provider goes down, it gathers the documentation and builds the ticket for a human to resolve offline.

#### Sensitive data, always protected

Card numbers and other sensitive data never get exposed to the model — the logic lives in code.

#### Actively expanding

Ailina (credit limit increases), blocked-account recovery, and elderly-user onboarding, already underway.

How It Works

## From inquiry to resolution, without exposing what's sensitive

01

#### The inquiry arrives at the Help Center

Through Intercom, with a photo, PDF, or receipt attached if needed.

02

#### Joaquina interprets the content

Reads the attached file and builds the case without asking the user to retype it.

03

#### Checks the relevant status system

Transactions, KYC, limits, or accounts — whatever the case needs — without exposing sensitive data.

04

#### Resolves or builds the ticket for a human

If the case needs offline intervention, it leaves the documentation ready for the team to pick up.

### Integrations & capabilities

Intercom

In-app Help Center where the team already operates

Multimedia

Photos, PDFs, and receipts interpreted automatically

Status APIs

Transactions, KYC, limits, and accounts live

Data security

Expansion

Ailina, account recovery, and elderly onboarding

Results

## More scenarios, the same data protection

The assistant doesn't just automate — it covers more ground without touching what it shouldn't.

#### Sustained automation

Most inquiries are resolved without escalating, with the same judgment the team already applied.

70-75% deflection

#### Broad coverage

From balance and cards to disputes and KYC, in one assistant.

20+ scenarios covered

#### Zero sensitive-data exposure

The code-based architecture never exposes to the model what it shouldn't see.

100% of sensitive data protected

### Ready to resolve more, without exposing what's sensitive?

Let's talk about how StudioChat can cover more financial scenarios while keeping sensitive data protected.

[Talk to us now](https://cal.com/studiochat)

Case study published with Takenos's authorization. Data as of July 2026.