NewsCommunityFleetDesignEngineering
media

Improving Data Discovery Process at Ayazona (Part 1)

Sep 22, 2021 • 3 MIN READ

We strongly believe in data-informed decision making. Whether we’re considering a big shift in our product strategy or we’re making a relatively quick decision about which item to add to one of our weekly household essential kits, data provides the foundation for our decision making. An insight is a conclusion drawn from data that can help influence decisions and drive change. To enable us to make faster, smarter decisions, we use a suite of internal products to accelerate the production and consumption of insights.

Understanding our users' needs

To kick things off, we are spending a considerable amount of time conducting user research to learn more about our low-income customers, their needs, and their specific pain points. This is a very interesting set of customers, as the discovery process is entirely different, and unique. In doing so, we can gain a better understanding of our customers' intent within the context of their needs, and use this understanding to drive product development.

Low-intent data discovery

Let's say we've just onboarded a new low-income household on the platform for the first time with a familiar profile within the context of budget bandwidth, so a team member opens up the available catalogue and puts together a list of basic common essentials. This is a low-intent discovery experience! We had a broad goal to put together the first essential kit and we didn’t have extremely strict requirements based on our last kit for this customer(for the most part we take into consideration the content of the previous kit).

Within the context of data discovery, a data scientist with low intent has a broad set of goals and might not be able to identify exactly what it is they’re looking for. For example, as a data scientist, I may want to:

  • find popular datasets used widely across the company,
  • find datasets that are relevant to the work my team is doing, and/or
  • find datasets that I might not be using, but I should know about.

High-intent data discovery

Picture this, we have been supplying a low-income household with the essential kit for 8 weeks, and have a clear understanding of their current nutrition intake, their favourite items, weekly income pattern, and every family member personalised needs. So we know exactly what to include in their next kit.

Looking at this in the data discovery context, a data scientist with high intent has a specific set of goals and can likely articulate exactly what they’re looking for. This mode of discovery is often more important to more tenured data scientists who may be familiar with some datasets but may be looking for something they haven’t used before that meets a certain set of criteria. For example, as a data scientist with high intent, I may want to:

  • find a dataset by its name,
  • find a dataset that contains a specific schema field,
  • find a dataset related to a particular topic,
  • find a relevant dataset located in a particular BigQuery project,
  • find a dataset that my colleague has used of which I can’t remember the name, and/or
  • find the top datasets that a team has used because I’m collaborating on a new project with them.

By understanding our customer’s intent, enabling knowledge exchange through team members, and by helping individual members get started with a dataset we’ve discovered, we’re significantly improving the data discovery experience for data scientists at Ayazona.

If you’re interested in helping us tackle similar problems or you’re a data scientist that’s looking to work at a company with a high impact, do not hesitate to reach out.

Ayazona’s nutrition & affordability recommendation engine. • Ayazona
Recommendation algorithms are steadily taking centre at the core of Ayazona products.

More from Data

Browse more