• Black Facebook Icon

©2019 Ixio

Our Work

At Ixio, we do it differently

Building and Mentoring In-house Data Science Teams

At Ixio, we offer practical data science mentorship, support and instruction designed to give your analysts and data scientists the confidence and knowledge to tackle complex problems on their own. 

Stories of Success

We worked with the newly formed analytics team of a digital media broadcaster to design and deliver online tutorials and face-to-face instruction structured around their own data sets and current projects. In a few months, their analysts had developed the skills and confidence that allowed them to uncover and report insights and complete repetitive tasks much more efficiently than they ever had before. They were also able to understand and build machine learning models and use them in their work.

 

We worked with a top 4 bank in Ghana  to structure their analytics and data science function and selected the right candidates to fill roles. We helped architect a data warehouse, assisted with banking metric definitions and consistency, automated daily, weekly and monthly bank reports, developed advanced modelling skills within the data team, set up a campaign testing framework and developed advanced data science solutions for customer targeting. We also created a GIS system for monitoring branch and ATM performance by location.

We are currently mentoring in-situ data scientists at two Banks in Rwanda. We help design, plan and guide their work and ensure it aligns to the bank's objectives. Our team is available online for any coding support required.

Executive Masterclasses

After recognising a gap between business leadership and technical data science work, we developed a half-day Executive Masterclass in collaboration with the African Leadership University.

The course covers fundamentals of data science and machine learning, from how it works right through to advanced business use cases employing machine learning and deep learning techniques. The course aims to empower executives with a deeper knowledge of data science work, so they can ensure success in data-related endeavours within their organisation.

Custom Data Science Solutions

We also help our clients develop custom data science solutions. ​Below are a selection of projects:

Price Elasticity Modelling: we built a forecast model that assesses demand as well as other factors that influence demand such as inflation, consumer price index and price. A number of pricing scenario forecasts, of incremental changes in price, were modelled. The model performed consistently well across all of the client’s products and achieved over 98% accuracy in forecasting demand with specific price changes. 

Default Modelling: we developed a model, incorporating a machine learning technique that predicts whether a customer will pay their account or not. The model is time sensitive and is used to make predictions monthly (as per customer payment due dates). A customer behavioural database was built on which we based our model. The model is over 90% accurate and was used in a payment promotion campaign which generated over $2M per month for the period of use.

Business Scoring: we developed a diagnostic system based on a scientific methodology that could be used to understand, rate, monitor and improve the competitiveness of Micro, Small or Medium Enterprises (MSMEs). The scoring system is based on the past performance and capabilities of the MSME as well as other factors that influence competitiveness. It is size and industry specific. The scoring system is delivered through a web application and creates an environment where the MSME is exposed to relevant opportunities and assisted to develop their own skills and competitiveness in order to grow.

Call Volume Forecasting: we developed a customized time series model of call volumes, taking into account historical calls as well as event data to forecast inbound calls into the support call centre. A customised web application to monitor call forecasting success was developed for the client. The model, when implemented, saved R6million per month in staffing costs.