Mobile Bank Back-Office Migration: As-Is, To-Be, Card Products

How do you migrate a mobile bank's back office when its documentation is missing?

Azati's analysts documented a Mexican mobile bank's back office as-is and to-be, drew BPMN and sequence diagrams, and supported card product migration to a new vendor platform. The client's work with the vendor drew on those materials. Explore our banking software engineering work.

Ask us about back-office migration
Outstaffed

business and systems analysts embedded in a bank's back-office and card product migration

As-is and to-be

full descriptions of the bank's back office, current state and target state

BPMN and sequence

diagrams of the main customer processes and of frontend and backend interaction

Why did a mobile bank need analysts before migrating?

The legacy system was not built for card products, and moving to a new platform needed a shared, written picture of how the bank actually worked.

The client is a Mexican fintech with a mobile bank, 100-200 employees, and individuals as its main customers. Retail credit cards are part of what it offers. Its legacy system was not designed for card products, so the bank was moving its backend business processes and its back office from one platform to another, owned by an outside vendor.

The back office covered customers, applications, cards, technical support, and administration. The client had tried to analyze it before, but its business analysts changed often, and nobody could carry the work from start to finish.

Azati joined as outstaffed analysts: a business analyst first, a systems analyst later, on a time-and-materials basis. According to the project team, the client chose Azati after interviews and found the rates attractive. This was the first engagement between the two companies.

Technologies used

C#
C#
PostgreSQL
PostgreSQL
BPMN
BPMN
UML
UML

The project at a glance

A back-office and backend migration to a new vendor platform for a Mexican mobile bank, with Azati providing business analysis first and systems analysis and integration work later.

ParameterDetail
ClientA Mexican fintech with a mobile bank, 100-200 employees
CustomersIndividuals, including retail credit card holders
Business painThe legacy system was not designed for card products, and the bank was moving to another backend platform
What movedBackend business processes, the back office, and card products
Azati roleOutstaffed business analyst, then systems analyst, on time and materials
MethodAgile and Kanban
Back-office modulesFive: customers, applications, cards, technical support, administration
Diagram typesBPMN process diagrams and sequence diagrams
StackC#, PostgreSQL

What made this back-office migration hard?

Five things got in the way: a system not built for cards, missing documentation, a shifting architecture, many departments with uneven requirements, and a client team whose analysts kept changing.

Challenge 01

A legacy system not built for card products

The old platform could not support the bank's card business, which is why the move happened at all:

  • Retail credit cards for individuals were part of the offer.
  • The legacy system was not designed for card products.
#1
Challenge 02

Documentation that was missing

The project lacked documentation, so the current state had to be established first:

  • The current state of the back office had to be established first.
  • The target state could be described after that baseline existed.
#2
Challenge 03

An architecture that changed several times

The target architecture was not fixed during the project:

  • Architecture decisions changed more than once during the project.
#3
Challenge 04

Many departments and inconsistent requirements

A back office touches many teams, and they did not always agree:

  • Communication across departments was complex.
  • Requirements from different departments were not always consistent.
#4
Challenge 05

A client team that kept changing analysts

Earlier attempts to describe the work stalled because ownership kept moving:

  • The client's business analysts changed often.
  • No one had been able to complete the analysis end to end.
#5

How did the analysts approach the migration?

They wrote down the current back office first, described the target second, and turned the main processes into diagrams that the client and its platform vendor could work from.

Reconstructing how the back office works today

Azati's business analyst studied the current state of the back office and produced a full as-is description of how it worked.

Describing the target state

A to-be description followed, covering how the back office should work after the move to the new platform.

Drawing the processes that matter most

BPMN and sequence diagrams covered the main business processes: what customers do in the mobile app, and how the frontend and backend interact.

Giving the vendor a shared reference

The client's interaction with the vendor that owns the new platform was built in part on these diagrams and descriptions.

Extending the work to card product migration

When a systems analyst joined, the work extended to integrations for migrating card products, on C# and PostgreSQL.

Moving a back office without current documentation?

Bring what you have, even if it is a few screenshots and a lot of tribal knowledge. We will tell you what an as-is and to-be analysis would need to cover.

Talk about your migration analysis

What did Azati deliver?

Four pieces of work: an as-is description, a to-be description, BPMN and sequence diagrams, and systems analysis with integration work for card product migration.

01

As-is description of the back office

Azati's business analyst studied the current state of the bank's back office and produced a full description of how it worked.

Key capabilities:
  • Analysis of the current back-office state
  • Full written as-is description
  • Written baseline for the migration
02

To-be description of the target back office

Azati also described how the back office should work after the move to the new vendor's platform.

Key capabilities:
  • Target-state description
  • Back-office processes on the new platform
  • Written to-be description
03

BPMN and sequence diagrams of the main processes

Azati prepared BPMN process diagrams for the main business processes a customer goes through in the mobile app, and UML-style sequence diagrams for how the frontend and backend interact.

Key capabilities:
  • BPMN diagrams of customer processes in the mobile app
  • Sequence diagrams of frontend and backend interaction
  • Material used in the client's work with the platform vendor
04

Systems analysis and integrations for card product migration

When a systems analyst joined, the work moved toward integrations for migrating retail credit card products off the legacy system, on C# and PostgreSQL.

Key capabilities:
  • Integration analysis for card product migration
  • C# development
  • PostgreSQL data storage
  • Retail credit card processes

What did the client end up with?

The client ended up with a written current state, a written target state, and diagrams of its main processes, which its work with the new platform's vendor drew on.

DeliverableWhat it covers
As-is descriptionHow the back office worked before the migration
To-be descriptionHow the back office should work on the new platform
BPMN diagramsMain business processes in the mobile app
Sequence diagramsFrontend and backend interaction
Integration analysisMigration of card products

What changed for the client?

It gave the client and the platform vendor written, diagrammed descriptions to base their work on.

The client did not share numeric results, so the effect is described in work products and in how they were used.

The current state was written down

Where documentation was missing, the as-is description became the record of how the back office worked.

The vendor interaction had a shared reference

The process of working with the vendor that owns the new platform was built in part on Azati's diagrams and descriptions.

The main customer processes were visible

BPMN diagrams made the customer's path through the mobile app explicit, and sequence diagrams did the same for frontend and backend.

Card product migration got dedicated analysis

A systems analyst joined to work on integrations for migrating retail credit cards.

Domain knowledge stayed on the team

The analysts came out of the project with a working understanding of a mobile bank's back office and its business processes.

What this project taught the team

Three lessons: write the current state before designing the target, treat a diagram as an agreement between teams, and expect conflicting inputs when many departments are involved.

Write down today's state first

With documentation missing, a written as-is description is the baseline for what the new platform has to replace.

A diagram is an agreement, not a drawing

A BPMN process model gives the client and the vendor one version of a process to point at.

Many departments mean conflicting inputs

Many departments and inconsistent requirements made conflicting inputs part of the work. This is how organization shapes system design in practice: a system reflects the communication of the teams around it.

What does a platform migration need first?

A platform migration needs four artifacts first: an as-is description, a to-be description, a process model, and an interaction model. This project produced all four.

ArtifactQuestion it answersFormat used here
As-is descriptionWhat does the current system actually do?Written description of the back office
To-be descriptionWhat should the new system do?Written description of the target state
Process modelWho does what, and in what order?BPMN diagrams
Interaction modelWhat do the app and the backend exchange at each step?Sequence diagrams

Team composition

Azati worked as an outstaffed analytical team embedded in the client's migration.

  • Business analyst: studied the back office, wrote the as-is and to-be descriptions, and prepared the BPMN and sequence diagrams.
  • Systems analyst and developer: joined later to work on integrations for card product migration on C# and PostgreSQL.

How was the engagement delivered?

Outstaffed analysts on time and materials

Azati joined the client's team through a time-and-materials engagement, starting with a business analyst and adding a systems analyst later.

Agile and Kanban

The engagement ran in Agile, and the business analysis work used a Kanban flow.

Inside a multi-department client

The project involved many client departments and several architecture changes, and it started without complete documentation.

Who this approach is relevant for

It fits banks and fintechs that are moving to a new backend platform and need the current and target states written down before the move.

  • Banks and fintechs migrating to a vendor's backend platform, with back-office processes that need to move with them
  • Lenders with card products on a legacy system that was never designed for them
  • Teams in banking and financial software that lack current documentation
  • Programs where a platform vendor needs diagrams and descriptions to work from
  • Companies that want stable analytical ownership after repeated analyst turnover
  • Teams that need integration engineering alongside the analysis
  • Programs that need QA of financial calculations once the migrated system is live

Frequently asked questions

If you're the one who has to justify paying for analysis before any code gets written on a platform migration, this FAQ is written for that conversation.

A back-office migration moves a bank's customer, application, card, and support processes to a new platform.

On this project, Azati described the current back office, described the target state, and diagrammed the main processes the new platform had to support.

Azati supplied business and systems analysts to a Mexican mobile bank migrating its back office to a new vendor's platform.

The work, which also covered the banking app backend, produced an as-is description, a to-be description, BPMN and sequence diagrams, and integration work for card product migration on C# and PostgreSQL.

An as-is description shows what the old system does today, which is the baseline for the new platform.

Here the documentation was missing, so the current state had to be written down before the target state could be described.

BPMN diagrams show who does what and in which order, and sequence diagrams show what the app and backend exchange.

Azati prepared both for the mobile bank's main customer processes, and the client's work with the new platform's vendor was built in part on them.

Plan the integrations between old and new systems around the gap: the legacy system was not designed for card products.

On this project, Azati's systems analysis and development work covered integrations for retail credit card migration, using C# and PostgreSQL.

Treat analysis as the way to create missing documentation, and write down the current state before designing the target.

Azati's business analyst produced a full as-is description of the back office, so the later to-be description and diagrams rested on written facts.

Yes, analysts can join a migration as outstaffed specialists on a time-and-materials basis, and this engagement worked that way.

Azati started with a business analyst, added a systems analyst later, and the team worked in Agile.

The architecture changed several times, many departments had to coordinate, and their requirements were not always consistent.

The documentation was also missing, so the current state of the back office had to be established before the target state could be described.

Planning a back-office or card product migration?

Tell us what you are moving and what is documented today. We will outline the as-is, to-be, and integration work it needs.

Book a call with our experts

Azati's related banking and fintech expertise

Explore our successful projects and see how Azati delivers measurable results for our clients.

Automation of Back Office Deposit Operations

Automation of Back Office Deposit Operations

60% reduction in manual deposit processing time
50% increase in back-office operational efficiency
98% system uptime achieved after load balancing and orchestration
  • Java
  • Spring
  • React
  • IBM MQ
  • IBM WebSphere
  • Jenkins

⚡ Pain Points We Tackled

One of Europe's largest state-owned banks was running an outdated deposit system that lacked key features and needed major architectural changes. Manual steps and fragmented components slowed back-office work, raised the risk of errors, and made regulatory compliance harder.

Our Approach

Azati worked in Scrum sprints with weekly iterations and two-week demos, collaborated closely with the in-house team, and delivered a modular solution that plugs into the existing banking ecosystem.

Applied Methods and Practices

  • Deposit operations automation: Streamlined key deposit and withdrawal workflows, reducing manual effort and operational time.
  • Pluggable architecture: A fully modular solution that integrates into the existing banking ecosystem.
  • Scaling and load balancing: Minimized downtime and optimized resource use.

Solution Features

  • 60% less manual processing time: Manual deposit processing time fell by 60%.
  • 50% higher back-office efficiency: Operational efficiency of the back office rose by 50%.
  • 98% uptime: Load balancing and orchestration kept the system available.
Online Solution to Transform Loan Processing in Banking Sector

Online Solution to Transform Loan Processing in Banking Sector

70% reduction in manual data entry across bank offices
60% faster loan approval process
65% time saved per loan application
  • Java
  • Spring Boot
  • Camunda
  • Apache Kafka
  • PostgreSQL
  • React

⚡ Pain Points We Tackled

A bank wanted to automate loan issuance, remove paperwork, and reduce manual errors while letting customers apply in malls and bank offices. Long counter waits, manual verification delays, and error-prone work across several internal systems made this hard to scale nationwide.

Our Approach

Azati ran Scrum with two-week sprints, first building a mobile app for real-time loan applications with QR codes, SMS confirmation, and digital signatures, then a unified bank-office system on a microservices architecture of 37 services.

Applied Methods and Practices

  • Mobile loan application: Real-time applications with QR codes, SMS confirmation, and digital signatures.
  • Unified bank-office system: Consolidated the bank's office tooling for mobile and web clients.
  • Microservices architecture: 37 services supporting deployment across all bank branches.

Solution Features

  • 70% less manual data entry: Automation and system consolidation cut repetitive employee tasks across bank offices.
  • 60% faster loan approvals: Real-time mobile app and DMS integration sped up approvals.
  • 65% time saved per application: Each loan application took far less time to process.
Banking Software Architecture Modernization

Banking Software Architecture Modernization

3-4x faster mobile app load
40-60% higher backend throughput
2-3x faster transaction history retrieval
  • D3
  • Oracle
  • WSO2
  • Java

⚡ Pain Points We Tackled

A financial institution's heavily loaded monolithic backend caused mobile app load times of around 15 seconds and limited system performance. The client needed to improve responsiveness and scalability without compromising security or regulatory requirements.

Our Approach

Azati redesigned the backend architecture by decomposing the monolith into independent microservices, optimizing SQL and data processing, and separating transaction history from the main mobile workflow. The resulting architecture improved performance, fault isolation, and scalability while providing a foundation for further development.

Applied Methods and Practices

  • Monolith-to-microservices transformation: Analyzed the existing backend and identified processing, data bus, and database bottlenecks. Azati designed independent services that could be scaled and maintained separately while reducing the impact of individual failures.
  • Architecture and infrastructure design: Evaluated infrastructure options against performance, security, and legislative requirements and designed an architecture that supported distributed processing and flexible resource allocation.
  • SQL and backend performance optimization: Refactored inefficient SQL patterns and reduced unnecessary post-processing to lower database load and improve backend response times.
  • Mobile experience optimization: Moved transaction history to a dedicated screen so the primary mobile interface could load independently of historical transaction data.
  • Performance validation and future readiness: Built and stress-tested an MVP of the redesigned architecture, validated load distribution, documented the solution, and prepared recommendations for future Spring Cloud integration.

Solution Features

  • 3-4x faster mobile app loading: Reducing load times from approximately 15 seconds to 4-6 seconds improved responsiveness of the core mobile experience.
  • 40-60% higher backend throughput: SQL and data-processing optimization increased backend throughput while reducing unnecessary database and server workload.
  • 2-3x faster transaction history retrieval: Separating transaction history from the main application flow improved the speed of accessing historical operations.
  • Greater system resilience: Independent microservices isolated failures and reduced the risk of a problem in one backend component affecting the entire system.
  • Scalable foundation for further modernization: The redesigned architecture supports horizontal scaling and provides documented integration guidance for future Spring Cloud adoption.
FinTech Payment Infrastructure Modernization

FinTech Payment Infrastructure Modernization

Lean engineering team supported by embedded platform expertise
Enterprise payment orchestration platform
Ongoing cloud platform engineering engagement
  • Cloudflare
  • AWS
  • Pulumi
  • GitHub Actions
  • Docker
  • PostgreSQL
  • Datadog
  • Grafana

⚡ Pain Points We Tackled

A UK fintech operated a payment orchestration platform with a lean engineering team responsible for product development, cloud infrastructure, and operational support. Infrastructure work and an extended search for senior DevOps expertise were diverting engineering leadership from platform architecture and payment capabilities.

Our Approach

Azati embedded platform engineering expertise into the client's existing organization, modernizing AWS infrastructure, Infrastructure as Code, CI/CD, monitoring, and operational processes without disrupting ongoing payment platform development.

Applied Methods and Practices

  • Embedded platform engineering: Integrated directly into the client's engineering organization to provide cloud operations, infrastructure support, and developer enablement alongside ongoing product development.
  • Infrastructure as Code and automation: Expanded Infrastructure as Code with Pulumi and cloud automation to make infrastructure provisioning and environment management more consistent and maintainable.
  • CI/CD modernization: Maintained and enhanced deployment pipelines across cloud environments, supporting more consistent releases as the payment platform evolved.
  • Cloud platform operations: Supported AWS infrastructure, networking, monitoring, and observability across the payment platform while maintaining operational continuity.
  • Secure platform operations: Implemented infrastructure changes within the client's established security model, including controlled access and authentication requirements.
  • Developer productivity enablement: Reduced the infrastructure support burden on product engineers and technical leadership, allowing internal teams to focus more on payment architecture and customer-facing capabilities.

Solution Features

  • Engineering leadership refocused on product architecture: With infrastructure operations and deployment support handled by Azati, technical leadership could dedicate more time to payment architecture, platform strategy, and new capabilities.
  • Additional platform engineering capacity without prolonged hiring: The client addressed a critical cloud engineering capacity gap after several months of unsuccessful recruitment for an experienced DevOps engineer.
  • More consistent infrastructure and delivery processes: Infrastructure automation and standardized deployment practices created a more maintainable foundation for evolving the payment platform.
  • Continued payment platform modernization: Cloud infrastructure, CI/CD, and platform engineering capabilities could evolve alongside new payment services without requiring the internal team to absorb the full operational workload.
  • Foundation for ongoing cloud evolution: The engagement established an embedded platform engineering model supporting continued modernization of the client's payment infrastructure and cloud environment.

Last updated

Got a job for Azati? Let’s talk business!

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

What's next?

  • 1. Tell Us Your Story
    Describe your project. We come back within one business day with team availability and a rough plan. NDA on request before the first call.
  • 2. Get Your Roadmap
    Receive a detailed proposal with scope, team composition, timeline, and costs tailored to your goals.
  • 3. Start Building
    Azati aligns on details, finalize terms, and launch your project with full transparency.