Case Studies

From government agencies and the international development sector,
to healthcare and finance, our portfolio of clients include organizations with enormous data sets. Take a look at some of the case studies from our portfolio.

Dashboarding tools for automobile leasing

There are plenty of dashboarding tools available in the market that have very similar functionalities and capabilities - and price tags. Choosing the correct vendor can be a daunting task. Several factors need to be balanced and various trade-offs need to be considered. Software features, usage complexity, scalability, deployment barriers and dependencies are some of the important aspects that factor into the final decision. We are consulted by a car leasing company to find the best tool for their use case. While all the ‘out of the box’ solutions performed several of the functions this company needed, none of them satisfied all requirements. A few workshops later, we produced detailed wireframes for a solution that addressed business needs directly. After receiving buy-in from senior management and the technology team, we launched into development. The software provided an overview of all key parts of the business: acquisition, services, car maintenance and dispute resolution. Performance metrics designed by business analysts and senior management worked with a live data feed to provide mission critical information. If the user required more information, we designed data manipulation functions that the business was comfortable with. The end result was an enormous success. Rather than poring through an endless stream of dashboards trying to create a disjointed overall picture, we cut through the noise and created a front end that senior business leaders could navigate and use easily. Deployment was a simple process, and a tool optimized on the organization’s tech meant that the business could start using the new tool without enormous investment into infrastructure - or licensing fees.

Better data modeling for the online betting industry

One of the greatest challenges data teams face is communicating their complex analyses and algorithms to business leaders. Without leadership buy-in, even the most excellently designed models do not reach the deployment stage. We were contacted by the principal data scientist at an online betting company who were struggling with this exact issue.The leadership was hesitant about decommissioning their tried and tested systems and modelling suites. Due to the highly competitive nature of the online betting world, customer retention is key. A sudden and drastic change in the organisation’s recommendation engines, marketing platform optimizers or CRM operations could cause a significant shock to the business. Lab environments, business stakeholders reasoned, can never completely mimic the dynamic and constantly shifting real world customer decision making processes.After a few workshops with the data science team we designed a tool that, instead of trying to explain the various moving parts of the modelling suite, actually allowed the user to see them in action. Every set of predictions for each phase of the modelling framework could be immediately seen and vetted by business leadership. Users were able to control and define any part of the entire modelling framework. The results could then be investigated by direct comparison to results from their existing models. Decision makers were able to understand how the new framework significantly enhanced the existing one.Illustrating a vast and complex suite of mathematical models as a cohesive system is extremely difficult. Decision makers see it as a ‘black box’ and allowing such solutions to replace methods that have helped them for years is an understable concern. An intuitive and visual tool to display cutting edge AI in action is a significantly more powerful (and informative) way to gain buy-in from stakeholders.

Empowering democratic institutions through better data analysis

Understanding the needs of the population in a country is essential for democratic institutions to thrive. The international development sector and civil society institutions rely heavily on surveys for gathering data needed to direct programming and funding. We were approached by an implementing organization for international donor agencies to help their team better understand and visualize data collected from the field by thousands of workers in 6 different countries. The needs assessment for which the implementer required help involved working a large data set which needed machine learning and natural language processing to correctly interpret the answers provided by an extremely diverse user base. The backend needed to be extremely robust while the frontend had to be simple and user-friendly so that field workers with minimal tech expertise could use the tool without any impact on the databases. In addition to being user friendly the tool also needed to be light weight so that it could be used by field staff in areas without access to broadband internet. Our team came up with a lightweight solution that could be used even on dialup internet. Our team designed the tool in a way that most of heavy lifting was done in the cloud so that the filed workers system used very little resources to run efficiently.
‍

We helped them. How can we help you?

Get in touch to let us know.
Contact US