Optimize search experiences
Elastic technology allows users of the company's creative applications and website to receive a consistent and highly advanced search experience.
Power machine learning across search
The company helps users with search, browsing, recommendations, and assistants through Elastic relevant results combined with its own artificial intelligence (AI) innovations.
Scale to meet demand from millions of users
Due to Elastic's resilient, scalable architecture, millions of customers can search billions of documents without compromising speed and reliability
Using machine learning to deliver sophisticated search experiences
Fortune 500 multimedia and creativity software company specializes in software solutions that aim to change the world through digital experiences. The company has millions of customers worldwide, from the largest brands to individual consumers, that use its applications to transform, inspire, and impact people through stunning visuals, innovative documents, and engaging digital experiences.
Whether as a creator or as a consumer, the ability to quickly connect people with content is a fundamental pillar of the creative journey. For this reason, the company has invested in providing fast and reliable browsing and recommendation search capabilities built into the majority of its applications.
However, this endeavor presents several challenges. Director of Engineering for the search platform says there are billions of assets in their cloud created with dozens of its tools. In addition, the assets themselves are stored in many different formats and file types such as video, photography, illustrations, and documents, each with their own characteristics.
To address this in the past, different product teams built search systems and features tailored to their individual product’s specific asset types and customer needs. But this resulted in an inconsistent experience when switching between products and led to operational inefficiencies with similar search features being developed and redeveloped by multiple teams.
Developing a consistent search experience
The team set out to build a consistent search infrastructure and experience throughout the organization to overcome these issues. The company was already using the Elastic Search Platform to power search in its first application. They saw the potential to leverage Elastic support for AI innovations to provide next-generation search and discovery capabilities across more of the product suite.
Elastic meets our most important criteria. It offers free and open technology that we can manage ourselves, deploy on the cloud, and scale to the needs of the business. It is also extremely efficient and cost effective, which supports our plan to deploy search across more products and more use cases.
This enables the company to offer consistent, sophisticated search and discovery features across dozens of products, tens of billions of assets, and its website, handling 600 million queries per day.
From a customer's online desktop app, a single search query can be entered, and personalized results across the entire suite of applications are returned. This capability alone can be invaluable when working with one asset across multiple tools as many creative workflows require.
Built in resilience, scalability, and speed
Additionally, Elastic gives the team confidence that search is resilient and scalable across the organization. "Reliability is one less thing that we need to worry about because Elastic is so dependable out of the box. It means we can add it to the heart of our search operations knowing that it will scale to the volume of demand that we expect today and in the future," says the Director of Engineering.
The speed of the Elastic platform is equally significant.
With Elastic, we can ingest and index data faster and manage billions of assets. From a latency perspective it also means that we can provide best-in-class response times. For example, image search now takes as little as milliseconds.
Architectural consistency is equally important from an operational and support perspective. Using its own open-source engineering expertise, the company can swiftly adapt Elastic to the fast-changing needs of users. They are also able to be more agile in terms of meeting and outperforming the user experience offered by competitors.
Supporting machine learning-driven search and discovery
This is vital at a time when AI and machine learning are driving a paradigm shift in both search and the management of media assets. With Elastic, the company can now expand its own homegrown AI search innovations and technologies to bring even more personalized experiences to users.
In addition to helping customers find relevant assets, we're empowering them to discover personalized content and better assisting them in their workflows. The power of Elastic’s machine learning capabilities enriches the user search experience in the areas of recommendations and discovery.
The Director of Engineering also highlights the importance of Elastic data to the machine learning process. "It's extremely valuable that Elastic processes data on ingestion so that it is automatically ready for our AI systems. There are also vector and embedding features that we can use as building blocks for our machine learning operations," he says.
There are many ways in which Elastic machine learning is behind their search features. Some examples include the ability to search images by color content and composition, such as object position and size. There is also the ability to pull up two images, highlight a different element in each, and receive a set of image results that include both selected elements.
The company is also innovating on top of Elastic machine learning to better understand "user intent." Apps can auto-fill search queries or make recommendations to the user based on previous behavior and preferences. Recent developments include a feature that automatically suggests edits to a photograph based on similar images in the user’s own catalog or the wider customer community.
Leading the future of creative collaboration
Looking ahead, the team is exploring the possibility of using Elastic to offer voice, audio, and video search. Additionally, moving beyond traditional text search, creatives may soon be able to find specific images or objects within a 3D model.
The Director of Engineering returns to the point that data in the Elastic Search Platform is machine learning-ready and is the key component driving all these search capabilities . They’re also looking forward to further collaborations with Elastic.
