Solving Data Problems With Design Thinking

DZone 's Guide to

Solving Data Problems With Design Thinking

Design thinking can be applied to data problems to solve them using an open, agile and user-centric approach.

· Big Data Zone ·
Free Resource

Daniel Lewington, Director of Product Design, and Manisha Jangra, UX Researcher, both at Refinitiv Labs, explain to Jo Stichbury how they apply an open, agile, and user-centric design thinking approach to solve data challenges successfully - and at speed.

What Exactly Is Design Thinking?

British industrial designer Tim Brown defines it as "a human-centered and collaborative approach to problem-solving, using a design mindset to solve complex problems."

Daniel Lewington and Manisha Jangra explain that Refinitiv Labs uses design thinking to work closely with customers to identify and tackle data problems. Through direct and iterative collaboration, the team can brainstorm potential solutions and get early feedback on prototypes, concepts, and ideas.

Core Values of Design Thinking

"We have three core values. We are collaborative, with customers, across teams and skillsets; we are user-centric because we put the human at the heart of what we do and don't use innovation for its own sake; and our approach is iterative, intending to fail fast to ensure constant improvement."

Lewington explains that he sees design thinking as a risk management tool.

"We limit our exposure to risk because we don't push a long way ahead in a project without getting feedback. This allows us to get to the heart of what the customer needs, and by putting a prototype in front of them regularly, we can change direction if we need to and get things right faster. Quickly iterating and refining our projects saves us time and money."

Refinitiv Trade Discovery Tool

The team recently used design thinking in a project to assist customers grappling with the Risk Factor Eligibility Test (RFET), introduced by the Fundamental Review of the Trading Book (FRTB). FRTB is a comprehensive suite of rules around market risk-related capital requirements for banks, developed by the Basel Committee on Banking Supervision.

To complete the RFET, banks need to source extensive amounts of trade and committed quote data for the previous 12 months, map this data to risk factors, and check whether there is sufficient activity to pass predefined liquidity thresholds.

Refinitiv's Trade Discovery product has the data in a machine-readable format to support customers with RFET calculations. However, several customers wanted additional tools to support the process of sourcing and discovering data, as well as to refine and calibrate their mapping rules.

The Trade Discovery Tool offers a solution to source information, define custom mapping rules, and calculate the pass rate for customers via a UI that allows them to fully explore and understand the data. The tool makes calculations on where they pass and fail, and visualizes how they can improve.

The team applied the Design Council's 'double diamond' approach to addressing this challenge, comprising four overlapping actions: Discover customer problems and needs; define them; develop potential solutions and deliver viable solutions.

There were ten cycles of development, feedback, and refinement, which fine-tuned the prototype to a well-received product.

The Trade Discovery Tool project was particularly successful because the stakeholders formed an advisory board to explain their needs and give input on prototypes, ultimately allowing the project to move fast and be completed within the planned 12 weeks.

Data Scientists' Need for Speed

However, it is not always so straightforward to gather requirements or get feedback from customers working in the fast-paced world of financial services. The team needed to adopt various strategies to invite collaboration on a project to build a data exploration tool.

The increasing volume and depth of data available within the financial services sector are matched by the growth in the number of data scientists hired to analyze it. Refinitiv Labs set out to build a resource to help data scientists easily find, evaluate, access, and use financial datasets, and establish quickly whether they are useful and relevant to their needs.

As part of the design thinking process, the team held in-person interviews with over 30 data scientists. They identified access to clean data early in the discovery process to validate ideas as a key requirement.

In addition to in-person interviews, the team gained further insight by referring back to the results of a Refinitiv Machine Learning survey of more than 450 data scientists and financial sector professionals.

An additional brief was the 'need for speed', because the value of the data is frequently time-dependent, and requires a fast response. To allow data scientists to get up and running quickly with unfamiliar datasets, the aggregator presents them with tools that data scientists are familiar with.

By design, the data exploration tool provides customers with datasets and services that can be immediately opened in a cloud-hosted Jupyter notebook, with examples to preview and run for rapid assessment.

Placing Customers at the Heart of Innovation

Initially, the design thinking approach challenged the team because their customers were so busy - part of the problem that the aggregator set out to solve.

However, the customers were soon incentivized to provide input. "Early collaboration showed great results: We could get it right earlier, and the users got it in their hands sooner," added Lewington.

Lewington explained that design thinking has taught the team to set aside their personal opinion of what their products should deliver.

"We listen to what our customers want. At the heart of what we do, there's a human with a problem. We may produce something that runs on a computer and uses machine learning, but we are motivated by the people behind the tool."

"It's not about thinking outside the box but extending the box. We want to stretch what our customers can do, but still keep them at the center, through meaningful innovation rather than hype and noise," added Jangra.

"Our customers are also gaining awareness of the advantages of design thinking. They see the benefit of giving insights, and likewise, see products shaped by insights from others in a similar position. Our team is open to collaboration to solve common data problems across the industry".

agile design, big data, data science, design, design thinking, machine learning, user-centric design

Published at DZone with permission of Jo Stichbury , DZone MVB. See the original article here.

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}