By Justin Hingorani, VP of Products, Duco
New technology is solving one of the oldest problems in computing. There’s long been an imbalance between the need for software solutions and the programming resources available to create and maintain them.
This problem has only intensified as technology has become more useful and ubiquitous in the corporate environment.
It’s not a problem you can solve with more developers alone. What’s needed is a change in the way firms utilise technology. One that bridges the gap between IT and the business, while removing work from IT departments overburdened by other priorities, such as security and maintenance.
Unblocking technology bottlenecks in financial services
Financial services firms have enormous and often complex application landscapes that require IT support. Many of these systems are hard-coded, often built on on-premise databases and are impossible for business users to configure themselves.
They must brief the IT team and then wait while their requests make their way through the development pipeline.
Understandably, all this can take a very long time: a survey Duco conducted with Financial Technologies Forum found that 13% of firms took up to two months to onboard a new reconciliation, 22% took three to six months and 9% took too long to even bother. This is why the spreadsheet has become the ubiquitous tool for handling data problems in financial services.
Why should you empower your users?
No-code is challenging this established order. It’s putting the power over data into the hands of the people who know it best: the business users. They can create their own solutions, increase productivity and cut out the repetitive manual aspects of their jobs.
This frees them up to focus on value-add tasks that help the business enhance the customer experience, create efficiencies, or improve competitiveness.
The business benefits hugely from this shift in capabilities and focus. Firms are constantly reacting to changing demands, such as regulatory changes – like the CFTC Rewrite and EMIR Refit, new financial products, or new data types. Reconfiguring processes running on legacy systems is a herculean task, especially as these systems often weren’t built for those purposes to begin with. They often have a fixed schema and are only built to accept data in a set format.
This places enormous pressure on IT and creates risk for the business. In the case of regulatory change, failing to meet new requirements can damage your reputation and make you liable to fines. No-code tools can help you avoid these issues. Processes that may have taken months to change or implement are updated in hours. This is particularly powerful in financial services, where change is often constant and at the last minute.
Your team will feel the benefit, too. Giving employees — particularly those usually involved in previously repetitive and valueless tasks — a better experience matters. Companies are competing
hard to attract talent (there were a record 1.3m vacancies in the UK during summer 2022) and against the so-called “Great Resignation” which became apparent during the pandemic.
But while no-code seems like a new way to help create business agility and improve the employee experience, it can trace its roots back at least 40 years.
A brief history of no-code tools
Renowned technologist James Martin wrote in 1982 that: ”The number of programmers available per computer is shrinking so fast that most computers in the future must be put to work at least in part without [them].”
Since then, the desktop environment, the internet, the smart phone and ever-increasing amounts of tailored software, have gradually put more creative technology in the hands of more people.
The term “citizen developer” was first used at the 2009 Gartner IT Symposium/Xpo in Orlando. Mendix, Salesforce, and Appian were given as examples of “quicker alternatives to traditional programming platforms”. Their low-code offerings, such as Salesforce Lightning, meant users could “accelerate app delivery by dramatically reducing the amount of hand-coding required.”
Low-code tools have many uses, but they are largely aimed towards developers as they still require some technical skills. No-code solutions developed from this foundation, but have a different goal: empower non-technical users to take control.
In 2021, Gartner predicted that, by 2024, 80% of all technology products and services would be built by citizen developers, a rise from 25% in 2014. The pandemic turbocharged this shift. According to Gartner, “the rapid expansion of cloud services, digital business initiatives and remote services opened the door for new possibilities in integrations and optimisation.”
Drag-and-drop interfaces are one of the most popular and useful ways of creating custom software without learning to code. They vastly simplify the creation of websites, web apps, mobile applications, and databases since users can build them in a graphical user interface (GUI) without code.
Great examples are the data quality dashboards used in financial services. These are built on business intelligence platforms (BIPs) like Tableau and PowerBI. They can be integrated with a technology solution via an API and then used to monitor a wide range of data quality issues, such as:
Exception trends over time
· Top causes of reconciliation breaks
· Sources of low quality data
· Status of unresolved exceptions
· These are a powerful way for the business to better understand the state of its data.
No-code data management with natural language programming
Drag-and-drop programming as it happens in a GUI is hugely beneficial to users creating their own workflows. However, it has limitations. For example, users can only draw upon the pre-coded elements to create their workflow. They can’t add new ones to solve the particular problems they are facing.
A more flexible way to create solutions comes via programming using words rather than a visual approach. This mix of linguistics, computer and data science is known as Natural-Language Programming (NLP).
These rules can be incredibly varied and powerful, from transforming data in certain fields to extracting text strings to setting thresholds.
Natural-language processing: the future of no-code solutions
Over the last year or so the boundaries between no-code and machine learning have become increasingly blurred. We’re heading for a world where even Natural Rule Language becomes unnecessary.
In this world, a machine learning-powered platform would ‘understand’ your meaning and code a rule for you.
The development of this future is underway and Natural Language User Interfaces (NLUIs) which execute programs are all around. You can see these in action when Gmail or Google Docs offers to auto-complete sentences with Smart Compose. And virtual assistants like Alexa and Siri can already interpret and carry out simple commands. But the latest innovations in this space go way beyond this.
GitHub Copilot is a pioneering tool that can generate code from natural language inputs, turning written English prompts into coding suggestions across dozens of programming languages. Meanwhile, Google has developed an autocomplete tool for use by its 10,000+ internal developers. 25% of single-line code suggestions were accepted, rising to 34% for multi-line code suggestions.
Utilising code-generation techniques like those above could allow end-users to build data transformation rules by describing their aims in plain English sentences. The computer would then interpret their meaning and write executable, tested code.
A pivotal moment for the way financial services firms operate
No-code is removing the large handoff points between operations and IT, empowering the former and removing burden on the latter.
It started with drag-and-drop, allowing users to get unrivalled access to insights regarding their data. Then natural language programming tools, like our Natural Rule Language, to create even complex data transformation rules in hours.
And it’s only going to get easier. One day users will be able to ask a computer to carry out complex, domain-specific requirements and have it respond. Simple.