How was Transferwise Developed?

TransferWise is a global technology company that offers a money transfer service around the world. Using TransferWise saves you money when you make purchases abroad while traveling, sending money to another country, or receiving and making international payments.

A combination of smart branding and advertising based on people’s resentment towards banks and some of the lowest exchange fees in the industry has made TransferWise one of the biggest international money transfer providers around the world.

Its growth has exploded since its launch, but who are the people behind one of the most valuable fintech start-up and the idea that started it all?

Meet the Creators of TransferWise

Kristo Käärmann

Kristo Käärmann was born in Estonia when it was still occupied by the Soviet Union. After ten years of growing up just like any Soviet kid at the time, his country became independent.

With its newfound independence, Estonia went through incredible growth, and embracing this new kind of Estonia, Kristo’s interest in mathematics and computers flourished. He studied mathematics, then he did a master’s in microbiology and the computation techniques around it.

At the same time, he got a job with PricewaterhouseCoopers where he was working as a management consultant, helping banks in Estonia that wanted to get a better grasp on their data.

But remember, Estonia was just newly independent as only 15 years have passed since the Soviet Union’s reign. That meant the oldest business you could find in the country was basically 15 years old and without much of a legacy.

Normally, consultants work with really old companies, helping them cope with change and deal with legacy. They show them advanced processes and newer, better ways to run their company.

Since these companies were young, they did not need the usual consultancy that older ones did. In their situation, they were facing different problems, like various complex processes and data challenges. As Kristo looked further into these problems, he realized that they are even deeper rooted around the world.

That was when in 2007 he moved to London and started working with Deloitte Consulting, doing a similar job as he did in Estonia, but with companies that were about 100 years older.

Taavet Hinrikus

Just like Kristo, Taavet is from Estonia, he was born only one year after his co-founder. The two are close in age, and they grew up in the same country at the same time, giving their future friendship and bond a strong base.

His professional background was focused on building internet businesses and product development and he started his first job in 1997 at Halo Interactive DDB as a Senior Engineer where he led the site operations and the development team.

After three years in this position, he continued his career at Equitygate.com/Hanashi IFC as a Project Manager, and here he was also responsible for the development team and their effort in building a new platform.

Later, in 2002 he became a Product Manager at Joltid where he was in charge of product development and the final launch of PeerEnabler.

however, the main determinative job, where he became a well-known name in his profession came afterward when he helped a company grow from a start-up to a widely used telecommunication application.

We are talking here about Skype. In fact, Taavet became the first-ever employee at Skype, working there as Director of Strategy from 2003 to 2008. He worked on developing their technology, business, and product.

How TransferWise started

In 2010, the two friends, Taavet Hinrikus and Kristo Käärmann created Transferwise because they found the international 5% bank transfer fee simply outrageous. They believed that money should flow freely.

So, basically, what started it all was frustration. Kristo got his salary in pounds but spent a lot of it on a mortgage in Estonia in euros. Taavet, on the other hand, while living in London got his salary in euros from Estonia.

Every month Taavet had to walk to a bank and transfer a few thousand euros from Estonia to England. Each month it took four days for the money to arrive. But taking the time to walk to the nack in person, and then wait some more for his paycheck to actually arrive were not the only issues.

When he checked the exchange rates and made the calculations of what he sent in euros compared to what he actually got in pounds, it did not add up. He realized that on top of the exchange rate, banks were also adding a hidden markup.

The world’s banking systems were not created with people and businesses without borders in mind so, instead of just exchanging through a bank, they came up with a simple scheme that would help them save money.

Each month they checked the day’s mid-market rate and found a fair exchange rate. Then, Taavet would put euros into Kristo’s bank account, and Kristo would deposit pounds into his friend’s account. This way neither had to pay their banks any exchange rate fees but got the currency they wanted.

This idea marked the foundation on which Transferwise will rise and revolutionize the way we transfer money worldwide. Because of their frustrations and that constant feeling they were being cheated by the system, the two friends made it their mission to charge as little as possible on international transfers.

TransferWise becomes Wise

As TransferWise took off, their services became more and more complex. Not only that they allowed for more affordable currency exchange, but they even included multi-currency exchanges, debit cards, and more.

A decade later, on Feb. 22nd of 2021, TransferWise became known as just Wise, to further showcase the simplicity their solution is offering to the market.

Currently, Wise is on average eight times cheaper than the leading UK traditional banks. And, they have over 10 million customers, who move over $6 billion through Wise every month., and the decision of using Wise instead of a bank saves their customers an approximate $1.5 billion each year. Looks like Taavet and Kristo had the right idea.

How does Wise work

Aside from their reasonable fees, Wise also differentiates itself by being built with users in mind. You can easily create your free account with Wise, where you can just sign up as an individual. And if you are a business owner, you can later add your business to your own Wise account.

Regardless, Wise is also mobile and accessible, so you can use and transfer money on your phone via their app, or on your laptop on their website.

Moreover, Wise is available for residents in most European countries, as well as the United States, Canada, Brasil, Australia, New Zealand, United Arab Emirates, India, Hong Kong, Japan, and Singapore.

And what’s even better, when you send money through them, it does not matter whether the recipient you are sending money to uses Wise or not, as long as they are residents of one of the formerly mentioned countries.

You can get your own Wise debit card for free, and pay with it anywhere, and there are no monthly fees or minimum amount you need to have on your card. In turn, you will at all times get the real exchange rate of the currency you pay with.

Why is Wise so much more affordable?

Traditional banks in general tend to operate in one country and in that country’s currency. They only have a small amount of other different currencies. So, when one of their customers wants to transfer money abroad, they exchange the amount at the Foreign Exchange, or FX, market rate. After this, they can send the money in foreign currency to the recipient’s account.

In other words, the fees banks charge are what it cost them to exchange the money based on the FX market rate, plus a large margin (~80%) of the cost. Currency exchange profits are then used to compensate for other parts of the bank account services that otherwise don’t make money.

Wise, on the other hand, uses a double-sided model. The company has bank accounts in every country they are operating in, so they do not actually transfer money abroad. Instead, they match users with each other, just like the two founders used to do for themselves in the old days.

This means that if someone wants to exchange euros for pounds, they get matched with someone who wants the opposite, exchanging pounds for euros. This way there is no actual currency exchange that would need to use FX, saving them the fees of operating in the market.

The technology behind Wise

Wise was built on the Grails framework, using a MySQL database, and the reasoning behind choosing Grails is that Kristo, who knew that banks have been using Java for fast development, was looking for the best Java/JVM based frameworks available.

He then ended up finding Grails and agreed it is a better, more efficient way of building on the Java platform. As a full-stack framework, Grail makes it easy to build on top of existing elements in Java, is using modern principles such as Don’t Repeat Yourself (DRY), and is a dynamic framework much like Django or Rails.

Not only that Kristo chose the technology, but he even wrote the initial first couple of thousand lines of their code himself.

They use MySQL for their database solution because it is a fast, relational database, it has gone through a lot of development and optimization, and new features are constantly added in the latest versions. MySQL and SQL databases, in general, are also known to be scalable which is why we often see them being used by companies like Netflix, YouTube, Facebook, Linkedin, and PayPal.

Another advantage of using MySQL is the fact that it is popular and widely known, so it is easy to find engineers who are familiar with it. However, Wise only uses the database as a storage engine, and all the business logic is put inside the application code.

Just like with any other start-up company, when choosing the technology Krtiso and Taavet’s main criteria was the capability of rapid prototyping. They wanted to get the product released as quickly as possible, this way they can get real user feedback before their commitment to a larger budget for full development.

This is a known practice within the technology industry, where the first iteration of the product they create is called a minimum viable product or an MVP. It means that there are enough features to attract early-adopters and validate the idea.

More importantly, releasing a simple MVP instead of a fully developed product helps with improvement in the early stages and learning what resonates with the target market and what does not, while keeping the financial investment low, allowing the startup to shift perspectives or even start over with minimum losses.

The bigger engineering challenges however came with their rapid growth. They needed to hire super-engineers, improve scalability and implement support for new markets and currencies.

As the amount of their data increased, they also needed to redesign different segments of their system. With the changes and advancements, it became obvious for their developers that to deal with their growth they needed to turn towards an architecture built around microservices.

Simply put, microservices are a different style of building the architecture of an application. Traditionally, features and functions were all written in the same codebase, and if something failed, everything would go down.

With this type of structure, however, instead of building features together in one codebase, it breaks the application apart into an assortment of, you guessed it, microservices that are:

• Owned by a small team
• Individually deployable
• Testable and maintainable
• Loosely coupled
• Formed around business capabilities

The microservice architecture allows fast, frequent, and dependable delivery of extensive and complex applications. Moreover, if one microservice fails, the chances of it affecting the rest of the website are close to zero. Think of them as plugins for WordPress.

Regardless, while this strategy solved numerous scaling problems, it also brought up new issues in the creation of simple reports and analysis of data.

To do these tasks, the team would have to connect to hundreds of different data stores, and development teams at Wise are free to choose what type of data store they want to use. As one can imagine, all of this created a way too complicated and impractical situation for analytics.

For this, they needed a new tool that had to be able to gather data from hundreds of sources into a centralized data storage so, data analysts can further do transformations and collect insights.

Their approach to finding the right tool was to look at existing data transportation products and try to find one that could satisfy the cases they faced, and also pass the meticulous requirements their security team had. So naturally, they looked at a number of traditional ETL tools, as well as commercial replication tools but they all fell short.

The problem with traditional ETL tools was that they did not support CDC and automatic schema changes well enough. Wise had dozens of databases that were changing each week and maintaining complicated ETL jobs just for this was unrealistic.

As for commercial replication tools, they offered interesting features. However, data would have to be sent to external servers which added unneeded extra complexity from a security and compliance perspective.

Then they found Kafka Streaming ETL which looked promising and seemed that it was a great solution for the company’s strong demand for real-time analytics. But, they ran into some difficulties with it, like the JDBC sink connector not performing well enough with some of the columnar analytical data stores.

After numerous months and dozens of different vendors later, they refocused on improving their internally developed tools.

Staying with their autonomous philosophy, they made a team responsible for every aspect of their data. They secured the extracted data in a central analytic data store. And soon after, they found the Singer specification that was created by Stitch and realized that they could get a working solution built on this.

So their analytics platform team, together with their own data analysts, made PipelineWise.

The project started out as an experiment and took several months to finish, but PipelineWise is now used by Wise each day to replicate hundreds of GB of data into their data warehouse from 120 microservices, 1500+ tables, and other external tools.

Conclusion

As we live in a world where companies operate in different countries and continents, they or their employees will possibly run into the same problems Taavet and Kristo did. So what we can learn from this is to keep an open mind, and try every solution possible until the right one comes along without sacrificing security, scalability, or user experience.

Wise did just that, and today they offer a great solution to a problem many have been having for years if not decades, and it is likely to grow even bigger than it is now with new customers constantly using it in more and more countries.

Like many successful companies out there, Wise was made out of frustration and grew out to be a company worth $5 billion. Without a question, what they offer was a wanted and much-needed service.

They save their users money, charging only an eighth of the price of some other money transfer possibilities. Their low fees are fixed and there are no hidden rate markups for exchange and they target the midmarket rate.

It is an obviously better choice than sending money abroad through banks since with banks, you will pay highly for the service and, especially if used frequently, lose a big amount of your money on just fees.

So if you have a frustration, as Kristo and Taavet had, and have an idea that can help you and many others like you fix that frustration, don’t hesitate to contact us. You don’t need to have development knowledge to make the world a little better, our experienced development team has you covered. So, let’s change the world together.

Read More Articles

https://wiredelta.com/how-was-revolut-developed/

https://wiredelta.com/how-was-paypal-developed/

https://wiredelta.com/how-was-google-developed/