Sunday, November 27, 2022

IBM: Watson Discovery Review

The modern enterprise environment generates new documents at an alarming rate, which makes analysis of information increasingly difficult. Researchers and customer service teams cannot waste time sifting through useless information — they need relevant data quickly.

IBM Watson Discovery empowers document search with artificial intelligence (AI) enhanced text analysis. Companies use the document search capabilities to index and search documents rapidly to enhance and power internal research, customer service, or chatbox services.

See below to learn all about where IBM Watson Discovery stands in the AI sector:

IBM Watson Discovery and the AI Market

Size estimates for the AI market vary depending upon its definition. The overall AI market includes both software and services and for 2021, estimates range from a high $328.34 billion from Fortune Business Insights down to $93.5 billion from Grand View Research.

AI software represents a more narrow subsection of the overall AI market. Gartner researchers forecast $62.5 billion for 2022 for this niche, which is relatively inline with the Markets and Markets’ estimate of $58.3 billion for 2020.

Compound annual growth rate (CAGR) estimates also vary but are independent of the definition. Fortune Business Insights and Gartner estimate CAGRs of 20%–21%, and Grand View Research and Markets and Markets estimate higher CAGRs of 38%–39%.

Key competitors to IBM in the general AI market include Advanced Micro Devices, AiCure, ARM Limited, Baidu, Enlitic, Google, Intel, Lifegraph, Microsoft, and Sensely. And Peerspot notes that the top competitors to IBM for indexing and search tools include Elastic Enterprise, OpenText, Microsoft, Lucidworks, Coveo, and Google.

Key Features

Key features of IBM Watson Discovery include:

  • Content mining
  • Custom dictionaries
  • Passage retrieval
  • Prebuilt connectors
  • Project-based user interface (UI)
  • Reading comprehension
  • Relevancy training
  • Smart document understanding
  • Support for up to 22 languages
  • Table identification and retrieval
  • IBM Watson Knowledge Studio (WKS) model integration
  • Data isolation, mutual authentication, and compute isolation for premium users

Key Benefits

Customers looking for AI-enhanced indexing and search can expect several benefits, such as:

Enhanced Context

Indexing and search tools powered by AI can categorize text, tables, and images. For internal documents, researchers can rapidly understand content and categories of information with full paragraph results returned to preserve context. The AI can also be trained to understand industry-specific language to understand context and provide even more accurate results.

In a customer service environment, Watson Discovery can report on trends, so organizations can apply analytics to understand customer requests and understand needs more deeply. These reports can be used to enhance the customer service experience and provide better information.

Enhanced Chatbots

Chatbots use natural language processing (NLP) to analyze customer input and search corporate resources for answers. Using AI-enhanced indexing and search tools increase the speed and accuracy of chatbots to provide more useful information on a more timely basis.

Rapid Analysis and Research

Companies create more documentation and receive more customer feedback. However, we have not become faster at reading and analyzing the ever-increasing flood of information that can overwhelm researchers.

IBM says Watson Discovery can cut research time by more than 75% and reduce knowledge worker time by 50%.

Use Cases

HSBC

HSBC, one of the top 10 banks in the world, wanted to use AI to search through decades of structured and unstructured data to identify investments with the highest probability of price appreciation. HSBC partnered with EquBot and deployed IBM’s Watson Discovery to help develop the world’s first AI-powered index fund.

“In today’s markets, investors need strategies that can keep up with the growing amount of data generated each day,” says Dave Odenath, head of quantitative investment solutions for the Americas at HSBC Global Banking and Markets.

“We are now able to offer clients solutions that not only keep up but also thrive in an increasingly complex world of data. AiPEX with Watson simulates a team of thousands of analysts and traders working around the clock to learn from millions of pieces of information and identify potential investment opportunities.”

Maricopa County Clerk of the Superior Court

The employees of the clerk’s office for the Superior Court for Maricopa county struggled to keep up with a daily volume of thousands of time-sensitive requests. Seeking to speed up employee response times, they deployed IBM Watson Assistant powered by IBM Watson Discovery to answer employee internal queries and create help desk tickets for complex issues.

Immediate success spurred the deployment of the solution as a self-help virtual agent, which in the first month of deployment handled 70% of citizen inquiries, saved 100 hours of employee time, and improved customer satisfaction.

“This is just one of the many innovations we have in mind,” says Aaron Judy, technology innovation strategist at the Clerk of the Superior Court in Maricopa County.

Adoption of Watson Discovery improved response time and also improved understanding of citizens’ needs. By analyzing citizen requests, the Clerk’s office could better understand what services to offer remotely, what actions to feature on their website, and other ways to improve customer service.

Woodside Energy

An independent oil and gas company in Australia, Woodside Energy, wanted to make the knowledge of its senior experts rapidly available to all engineers — even after retirement. Woodside needed a tool to analyze 30 years and 600,000 pages of documents, reports, and correspondence as well as to help engineers stop spending 80% of their time researching solutions.

Woodside selected Watson Discovery to take advantage of the machine learning model that could analyze natural language and be trained for specific industries. Over 80% of employees adopted Watson for day-to-day work and now only spend 20% of their time researching solutions.

Woodside estimates that the company is saving over $10 million dollars worth of engineering time.

“You have to spend a bit of time training Watson before it will actually work, but it’s been really positive in the operations area,” says Caitlin Bushell, a graduate process engineer for Woodside.

“It’s helped our engineers get up to speed very quickly on what has already been done and how they were managed in the past.”

Differentiators

Several companies compete with IBM to offer indexing and search solutions. Customers selecting Watson Discovery do so because of key differentiators such as:

Brand and Product Reputation

IBM remains one of the most recognizable brand names in any corner of the computing industry, which helps to add value beyond the technical capabilities of IBM’s solutions.

IBM Watson also helped pioneer AI computing, and IBM Watson tools remain recognized market leaders in many categories by industry analysts including:

  • Leader in Gartner’s “Magic Quadrant” for insight engines 2021
  • Leader in the Forrester “Wave” for AI-based text analytics platforms (document focused), Q2 2020, receiving the highest score for strategy and market presence
  • Leader in the Forrester “Wave” for AI-based text analytics (people focused) Q2 2020, receiving the highest score for strategy and market presence

Flexible Deployment

All competitors in the AI-enhanced indexing and search market offer cloud-based solutions. IBM Watson Discovery is one of the few that can be deployed to multicloud or private cloud environments through the IBM Cloud Pak for Data.

Natural Language Understanding

IBM Watson’s natural language text analytics quickly analyzes and understands internal documents or client requests. Incorporating natural language features into search or a chatbox allows for quick and accurate responses.

Watson AI Tools

Watson Discovery is one tool in the suite of AI-enhanced options available under the IBM Watson umbrella. Watson Discovery integrates tightly and compliments other tools, such as Watson Assistant (virtual agent) to provide greater value.

IBM’s WKS is another tool available within Watson Discovery that allows users to teach the AI the language of the local dataset with custom models. These code-free models identify entities and relationships unique to an industry or company.

User Reviews

Review site Rating
Gartner Peer Insights 4.4 out of 5
TrustRadius 7.2 out of 10
PeerSpot 3.8 out of 5

Pricing

IBM offers Watson Discovery with four levels of plans: Plus, Enterprise, Premium, and IBM Cloud Pak for Data. Plus is available with a 30-day trial, and pricing starts at $500 per month for up to 10,000 documents and 10,000 queries per month.

The Enterprise level increases the quantities to 100,000 documents and 100,000 queries per month starting at $5,000, and IBM also provides a cost estimator to help clients understand how volume might affect pricing. IBM does not publish rates for the Premium or Cloud Pak levels of access.

Conclusions

Companies seeking to analyze and search through large datasets can rapidly accelerate progress using AI-enhanced tools. IBM’s industry-recognized Watson Discovery solution digests structured and unstructured data to help locate key information more quickly.

With the ability to integrate with other tools in the IBM Watson family and the backing of IBM’s strong brand, any enterprise considering AI indexing and search solutions should include IBM’s Watson Discovery on their short list for testing.

Similar articles

Latest Articles