Important aspects to consider when ordering a machine learning component

Machine learning has gained wide popularity and every day you can hear about a new way of its application. But such popularity has a negative side. In our practice, we often face a misestimation of its capabilities and the complexities of its implementation. To organize an effective development process, it is important not only to understand the needs of the customer but also to form the right customer expectations. This is the main reason why we’ve created this series of articles..

Registration and Onboarding

Ease-of-use applications have a great opportunity to succeed. As such, intuitive registration and onboarding processes are very important. However, finance applications usually require lots of data to verify starting from simple verification of a client’s bank account or the basic client’s data collecting and verification to usage of biometric technology. Depending on the complexity of the app and the scope of features the registration and onboarding development takes from 250-400 work-hours.

To open next 7 Features and Estimates of PFM App leave your email
Be the first one who gets new blog posts and updates!
Bank account aggregation

The app can assemble all types of accounts that a user has: credit and cards, loans, and other bank details. This way you’ll provide a one-stop solution for money management and contain all kinds of confidential data. That’s why security is a top priority and you’ll have to comply with a range of security standards, such as PCI DSS (Payment Card Industry Data Security Standard) and GDPR (General Data Protection Regulation). Approximately development of this feature takes 120-200 hours. 

Expense tracking

The app keeps track of the user’s finances in real-time, and there’s no need to switch to other finance management apps. We all have expenses that consistently eat our budget, for example, coffee, cigarettes, clothes, etc. The tracker can also trace all these expenses, notify you how much money you spend, and set a limit on them. It takes 160-300 man-hours.

Tax expense tracker

Upon any purchase with any of the user’s debit/credit cards the app detects the purchase and defines it as a tax expense or not. Thus a user has a list of spendings that can be downloaded for tax purposes Quarterly or Annually. This feature may take 220-300 hours due to different tax policies of different countries. 

Smart balance. Predictive model

The predictive model collects statistical data and analyses the user’s regular expenses using a machine learning technique to predict the future expenses. As a result the model provides a “smart” estimate of the balance after subtracting projected future spending, so a client knows what sum of money he has when the regular expenses are done. Implementation of this feature takes about 120-400 hours.

Budgeting and saving goals

Users can compare their entire monthly budget and actual spend across different categories (e.g. clothing, groceries, cinema). The app also allows setting up saving goals for something users have been dreaming of. This feature will cost you from 350-700 hours depending on the app complexity in general.


To provide the customer with full control of his personal finance the application may be equipped with the payments functionality. So there is no need to go to your bank application. A user can pay right in this app. This core feature takes 260-320 hours.


This cool feature allows buying stocks according to a particular investment strategy. A user takes a quiz on their goals and financial opportunities. On the basis of this quiz and analytical data taken from the user’s spendings and revenue, the app suggests different investment portfolios: conservative, moderate and aggressive portfolios. The app may connect to a stockbroker via an online API and automatically buy stocks per defined strategy. This feature can be supplemented with subfeatures that can essentially increase the development costs. Roughly it takes 1000-1500 hours.

Credit utilization

Credit Utilization Ratio (CUR) indicates how much the user currently owns on all credit cards compared to the cards’ limits. CUR = Outstanding Credit Card Balance / Credit Limit * 100. CUR is a key factor in making good spending decisions especially for those who make a lot of transactions per day. Ballpark estimate for this feature is 300-400 work-hours.

Who Develops a Personal Finance App

In-house team

The first variant that comes to mind is to build an in-house team. Yet, it’s not cost-effective: extended time to market, substantial hiring and office costs, the need to keep in-house specialists even without sufficient workload. What’s more, you need to hire an experienced CTO to manage staff. Your own in-house development team is the most expensive option.

App development agencies

Personal finance app development with an agency allows you to be sure in trustworthy transparent relations, quality of all deliverables, fixed schedule, all services (or at least the majority) in one place, and effective project management. This option is cheaper than setting an in-house team and it’s more reliable and safe than working with freelancers, for example.


Some freelancers can cost less, but the risks often outweigh this benefit. You may be dissatisfied with the service quality problem, flexible working times of freelance developers, lack of transparency and accountability, difficulty to manage several freelancers for several services. Any feature-packed project requires a substantial team of developers, DevOps and QA engineers, and designers. It is very hard to gather and manage such a group of freelancers. And one even very expensive freelance won’t replace the team. There is the risk to ruin the app and after all turn to an agency to fix it or rebuild from scratch.

Developers’ Location

App developers’ location is another major factor that determines the cost of a personal finance management app. There are several most popular areas where app developers are concentrated: the USA, Eastern Europe, and Asia

The most expensive option is to work with the US based developers. US and Western European developers charge around $80 – 150 per hour for the services. Asian countries (e.g., India and Pakistan) have the lowest development cost ($15 – $65 per hour). The rates of Eastern European companies are in the middle ($30 – 50 per hour). Currently, the latter variant is the most popular. The low cost of Asian Software development services often equals low quality and slow development processes. There are hundreds of cases when client companies had to switch to Eastern European developers (e.g., Belarus, Russia) to takeover a delayed or a failed project from a low cost vendor. That’s why a lot of Eastern European development companies, as well as Itexus, have a dedicated option for the projects with low quality development experience – Project Rescue Services. 

Belarus is becoming a new silicon valley for software development companies. Good education, growing IT community, european culture, geographical proximity and competitive rates make this region an excellent choice for outsourcing development offering one of the best value/money ratio on the market. 


If you plan to develop a mobile app you will need to make a choice between a native and a cross-platform app. 

Some Facts about Native Platforms

  • Recommended by Apple and Google
  • Best user experience and performance
  • Ability to directly use all device resources which results in fast load times easy access to GPS or Camera
  • Get regular software updates if they are designed properly and properly synchronized with the OS

It makes sense to have apps on two native platforms: iOS and  Android. You can start off with launching an MVP on one of these platforms, check the concept and then launch the second one. Mind that the quantity of Android phone users prevail over iPhone ones’. But companies often target iPhone users first as one of the biggest paying market segment.

Back to top

Any questions?

Drop us a line.