Document repositories continue to balloon, yet organizations need to rapidly search their contents to support customer service, technical support, and other time-sensitive business needs. In addition to speed, modern solutions must be able to handle many types of documents and extract information from images, audio, video, and the metadata of all files.
Organizations turn to scalable cloud resources such as Microsoft’s Azure Cognitive Search to apply artificial intelligence (AI), optical character recognition, and other features to create robust search solutions. These solutions deliver scalable speed and accuracy at a price that also scales with use.
See below to learn all about where Microsoft Azure Cognitive Search stands in the AI sector:
Microsoft and the AI market
Fortune Business Insights estimates the entire AI market to be $328 billion with a compound annual growth (CAGR) of 20%. Grand View Research estimates a lower market size of $93 billion and a higher CAGR of 38%.
Microsoft’s Azure Cognitive Search product competes in the software sub-section of the overall AI market. In 2020, Markets and Markets’ estimated the AI software market to reach $58 billion with a CAGR of 39%. By 2022, Gartner researchers forecast a market size of $62 billion and lower CAGR to 21%.
Microsoft Azure Cognitive Search competes in the AI-powered indexing and search category. Key competitors to Microsoft in the general AI market include AWS, Google, and IBM. In the indexing and search market, some top competitors to Microsoft include Elastic Enterprise, IBM, and Google.
Microsoft Azure Cognitive Search key features
- Fully managed search-as-a-service
- Geospatial search
- Filtering and faceting capabilities
- Quickstart templates
- Community and professional support
- AI-enhanced features, such as optical character recognition (OCR), key-phrase extraction, named entity recognition, Semantic Search that understands user intent, and search results ranked by relevance
- Built-in security and compliance features that meet a broad set of international and industry-specific compliance standards, including a fully encrypted indexing pipeline, support for virtual networks, the ability to control per-user access with security filters, and multilayer Microsoft security across infrastructure
Companies choose to pursue AI-enhanced search with Microsoft Azure Cognitive Search, because they seek the following key benefits:
Private data centers are fixed in size. If they are built to be big enough to meet peak needs, there are many expensive resources sitting idle when demand slacks. Adopting cloud resources permits flexible usage that scales up or down as needed, and customers only pay for the resources consumed.
Better quality search
AI-powered search can recognize images, people, locations, organizations, and phrases to enhance indexes and convert unstructured data into searchable content. In addition, AI-enhanced OCR, translation, and key phrase extraction can be selected with a checkbox to deliver better search results across a broad range of documents.
As the number and type of documents to be searched continue to increase, normal search becomes slower, especially when run on local data centers with fixed sizes. Shifting to the cloud permits additional resources to be applied to make search faster, and AI-powered indexes and searching produces even faster results.
Audioburst seeks to organize the world’s audio content to create a search engine for podcasts, newscasts, and interviews.
However, they could not afford to create an in-house solution to manage audio conversion, search indexing, or to service search requests.
“As a startup, we wouldn’t have been able to scale Audioburst and achieve our mission of organizing the world’s audio content without a cloud platform like Azure and advanced services like Azure Cognitive Search,” says Gal Klein, co-founder and CTO, Audioburst.
“We use it as our search engine to automatically index our content, make it more searchable, and offer recommendations.”
Ecolab’s technicians must support thousands of different operational, safety, sustainability, and hygiene solutions for 40 industries worldwide.
When operating from a client’s site, the technicians needed a quick and accurate way to search Ecolab resources to order parts or obtain specific information needed for repairs or technical support.
The Ecolab Virtual Assistant (EVA) was developed using Azure’s Cognitive Search and Azure Cognitive Services’ Language Understanding (LUIS) to handle natural language requests in multiple languages.
“Whether a field agent says, ‘Find me X product,’ ‘Look up product X,’ or ‘List assemblies of dishmachine X,’ EVA can understand and respond to the query intelligently through a combination of LUIS and the Azure Cognitive Search algorithm,” says Masaood Yunus, director of M&A, innovation, and enterprise architecture, Ecolab.
“And if EVA can’t find exactly what they’re looking for, it will provide alternative options for products, parts, or support documents.”
GE Aviation’s Digital Group determined that creating digital twins of a customer’s airplane engine could improve fuel efficiency, maintenance cost, and flight readiness of the planes in general.
However, to model the digital twin, they needed to ingest millions of pieces of paper that varied from customer to customer and frequently contained handwritten notes.
GE utilized Azure’s Cognitive Search to OCR and digest the text of those scanned records and power indexing and search with AI.
“In aviation, we think differently about what an asset is and what that means. An asset can be anything — like a single blade within a module within an engine attached to an aircraft,” says Jon Dunsdon, CTO, GE Aviation’s Digital Group.
“We can trace the history of each asset at the component level through to the whole airframe. We have a multilevel history and asset tracking capability to know when assets are removed and when they come together in different configurations.”
Potential customers select Microsoft’s Azure Cognitive Search solution because of the following differentiators:
Azure Cognitive Search deploys on the Azure platform and integrates tightly with other Azure-based servers and services. As one of the top cloud solutions, Azure delivers competitive pricing, strong capabilities, and robust support from both the Azure Community and Azure’s professional support team.
Deploying Azure Cognitive Search bypasses the operational overhead consumed by debugging index corruption, monitoring service availability, or manually scaling services during traffic fluctuations. Azure’s fully managed cloud search service deploys through a REST API or .NET SDK as a plug-in solution.
Azure’s Cognitive Search supports multiple layers of security to protect the data flow, such as TLS 1.2 encrypted data flow, inbound rules for IP firewalls, and private endpoints to fully shield the instance of the cloud application from the internet. It also connects to Azure Key Vault, Azure Private Link, and Azure Active Directory’s AD Login for additional security options.
Moreover, the security can be granuarly set to allow per-user control over search results. In other words, an organization can permit sensitive documents to only be located by people with sufficient security clearance. While more difficult, controlling access to indexes can also be implemented.
As a market leader in both on-premises and cloud computing solutions, the Microsoft and Azure brands are typically trusted and well known. Customers don’t worry about incompatibility issues as much for systems that have all been developed by a leader in software development and sales. Microsoft also provides partner and in-house resources worldwide that can provide customer support in local languages and time zones.
|Gartner Peer Insights||4.3 out of 5|
|TrustRadius||8 out of 10|
|G2||3.9 out of 5|
|PeerSpot||3.5 out of 5|
The pricing for Microsoft’s Azure Cognitive Search depends upon several factors:
- The Azure region in which it is hosted
- The hourly or monthly commitment
- The level of service needed
The service level defines the storage, number of indexes, scale-out limits, image extraction limits, and custom lookup limits.
For example, in the central U.S., the pricing ranges between the Free tier and the most expensive tier of $7.677 per hour or $5,604.21 per month. The lowest paid level is $73.73 per month or $0.101 per hour.
The pricing does not vary significantly for hosting in much of the world, but government and Japanese prices tend to be slightly higher. For example, U.S. government for the state of Arizona pricing ranges between free and $9.597 per hour or $7,005.263 per month.
Custom extraction, AI-powered content augmentation, and Azure bundle pricing are also available, but some of these require direct contact with Microsoft sales specialists. Microsoft provides a pricing calculator for estimates of multiple Azure products.
For enterprise-class search requirements, the cloud offers the cost-effective options for both scale and performance. Using an AI-powered search tool, such as Microsoft’s Azure Cognitive Search compounds the capabilities of cloud-based resources to deliver better and faster results.
Azure Cognitive Search should be on any organization’s evaluation list strictly for the powerful technical capabilities, but the other capabilities and differentiated features should make evaluation of this tool a priority.
For those who remain unsure, Microsoft’s free tier also allows potential customers to test out capabilities and work out integration difficulties prior to making a commitment.