By Chris Comeau, Principal Systems Design Architect Engineer
We’ve learned a bit about Data Mesh recently, and Octo offers its data mesh solution with processes and tools to address data management challenges. Let’s flip the script now and talk about how we can improve customer digital experiences in our applications by “meshing” all this data into user friendly search experiences — in other words, “Just give me what I’m looking for” in two clicks or less.
Challenges in UX and search
When it comes to searching big data that spans over multiple data sources, meshes, and data lakes, one of the biggest challenges in user experience (UX) is providing fast results that make sense to the user. Users need results that can match their phrases, misspellings, terms, synonyms, or just about anything they can type into a search field on a website. In an ideal web application, users could literally just be presented with an empty box, that says, “Type anything you like, and we’ll find it for you, within reason.” While a tool like this might be reminiscent of something used for a Google search, it would be specifically designed for the customer’s data and data provided by their affiliates. However, most search tools are unable to meet this objective, much to the dissatisfaction of the end user. This is where Octo has pushed UX beyond the user interface look and feel to tailor search experiences that seem to read users’ minds.
Taking search to the next level
At Octo, in support of one our biggest federal customers, we are improving the digital experience by focusing on delivering results by providing solutions that enable customers, partners, and employees to find what they’re looking for quickly. We are working toward solutions that solve significant pain points where there are big datasets spanned over multiple external data sources. We index this data into fast, performant, smart, and scalable search endpoints across millions of records. For the application users, this means they can simply visit our website, type in a box, get suggestions and find what they’re looking for — fast.
Imagine you are a user. You have a search engine that interprets as you type, not one that conducts wildcard searches, but a service that acts like a virtual assistant to suggest matching items. What’s more, these suggestions can be analyzed, researched, and trained with machine learning behind the scenes. Your results can be scored and boosted for relevancy and matched with synonym repositories that guide you quickly in finding an exact match.
On the development side, we can then identify trends, through which we may be able to pinpoint significant changes in search term popularity and incorporate the ones that are more relevant and frequently related to what users may type in a search. Not only that, but we can include faceted searches that can be used for quick user filtering, much like our favorite eCommerce sites, but without the dark arts of persuasive bait and switches. Our customers can monitor these searches, analyze this search data, and learn just how their users are applying their products.
These types of analytics run much deeper than client side, browser analytics which can be blocked. The solution provides an endless feedback loop for improving the digital experience across our applications.
The source of great search
Providing focused search results begins with the data mesh process, and it continues to refine the search user experience with fast lookups, popular suggestions, and aggregated keywords that support trends and boosted relevancy. In this sense, Octo is building search engines that continuously evolve and improve with each and every user click. It’s not just a fascinating concept. It’s a functional solution enabling federal customers to solve complicated and urgent problems because end users now have the right data right when they need it. And it’s positively impacting how they then in turn serve the nation’s public.
Are you ready to take search to the next level and meet your federal missions faster? Email firstname.lastname@example.org and ask about our search solutions.