AI & AnalyticsTechnologies

How to choose the best AI-powered analytics software in 2019

For almost as long as there has been business, there has been business analytics. Frederick W. Taylor is often credited with introducing business analytics in the late 1800s. Artificial intelligence hasn’t been around for quite so long but the start of modern AI can be traced back to the 1950s.

Artificial intelligence (AI) software is developing at a rapid pace. The AI market is estimated to be almost $90 billion by 2025, up from just over $3 billion in 2016. Corporations are using machine algorithms, predictive modelling and advanced analytics to identify trends and insights concealed in vast reams of data, allowing them to make faster decisions and keep ahead of the competition.

We’re arguably still in the “early adopter” cycle for the combination of AI and sophisticated analytics, but it won’t be long before the best AI-powered analytics software will simply become an essential part of a successful business.

There are is great deal of synergy between AI and analytics that serves to multiple their benefits so that together, they enable you to streamline processes, optimize customer engagement, automate operations, and improve decision making. Sometime known as cognitive computing, OpenText™ calls this “AI-powered analytics.”

What is AI-powered analytics software?

As the name suggests, AI-powered analytics is where artificial intelligence meets analytics. We can break this down a little further to identify three separate but integrated elements:

Artificial intelligence

Artificial intelligence means the ability of software to perform human-like functions such as perceiving, reasoning, learning, interacting with the environment, problem solving and exercising creativity to form plans, make decisions and achieve goals. There are a number of artificial intelligence technologies.

Machine Learning

Machine learning is the most important type of AI. Machine learning is a method of data analysis where machines use algorithms to detect patterns, learn to make predictions and make recommendations by finding hidden insights in your data without being explicitly programmed where to look. Machine learning algorithms are ‘trained’ by giving them access to more and more related data sets. The more data they have to learn from, the more accurate the predictions.

Business Intelligence (BI)

BI refer to tools, technologies and services that collect, analyze and present a company’s raw data as actionable information. BI software will include capabilities such as data mining, online analytics processing, search and querying, and reports and dashboards. Business intelligence analytics goes far beyond simple reporting–in areas such as sales figures, financial reporting or regulatory compliance–to provide a fuller and richer picture of the state of your business.

Many of the features within the best BI tools are also core to AI-powered analytics tools, especially the reporting and dashboarding features that provide information and insight in a way that is easy to understand for all levels of users. Humans are visual creatures. We understand and learn new information more quickly and easily when it is presented in visually appealing graphic images or recognizable patterns rather than in columns of data in a spreadsheet or database.

Advanced analytics

BI delivers descriptive analytics where historical data is used to understand what has already happened. Advanced analytics tools apply data mining, predictive algorithms and predictive modelling to existing data to predict future outcomes. This is often called predictive analytics. It is perfectly possible to conduct predictive analytics without computing technology–as long as you have a good deal of knowledge in data science, a lot of time and not a great deal of data.

However, organizations are faced with Big Data. They have an exploding volume of structured and unstructured data from a wide range of sources being created at ever more alarming rates. Handling big data is a huge challenge but also a massive opportunity because within that information is the insight that can drive business success. For example, the data gathered from today’s connected vehicle is helping automotive manufacturers create new data-driven revenue streams in areas such as Predictive Maintenance.

Predictive analytics, today, commonly includes many techniques including data mining, statistics, predictive modelling and machine learning that are easily able to cope with big data. It identifies the patterns and trends in historical data to identify business opportunities and risks in the future.

AI-powered analytics software includes a combination of these three technologies areas. It’s easy to confuse this with big data analytics. In fact, it’s best to consider big data analytics a subset of AI-powered analytics. (Look for a separate blog on big data analytics in the future.)

 7 core capabilities of AI-powered analytics

Things to consider when choosing your AI-powered analytics software:

1. Big data & big content

It’s essential that your AI-powered analytics tool can work with both structured and unstructured data. In addition to structured data (i.e. numbers and short menu selections) from enterprise applications, such as your ERP system, and databases, your data increasingly comes from unstructured sources such as enterprise file shares, email and social media. Today, an estimated 80% of data is unstructured held in content such as documents, PDFs, web pages, mobile content and videos. The best AI-powered analytics software contains both data mining tools and text mining tools to unlock the value of information locked in big content.

2. Data preparation

The AI-powered analytics software you choose must be able to connect to, blend and analyze data from any source. It needs to be able to identify and extract structured and unstructured data from data repositories both inside and outside the organization.

When social news and online new outlets are as important for a customer sentiment analysis tool as the data from internal sales systems, you have to be able to bring all data sets together within a data lake where it can be cleansed and standardized to enable data analytics tools to be applied to it. The best AI-powered analytics software allows data to be joined, extended and enriched in an easy to use visual drag-and-drop environment.

3. Algorithms and models

There are many different analytical, statistical and machine learning algorithms. A common misunderstanding is that data scientists create the algorithms. While this is occasionally true, the data scientist is far more likely to be using the algorithms that come pre-packaged within their chosen analytics solution to develop predictive models for the business area that requires analysis.

The best AI-powered analytics tools include a wide variety of the most common algorithms–regression, classification, decision trees and Bayesian algorithms–to enable data scientists to quickly begin to develop the decision-making models that the business requires.

4. Scalability

In the world of big data, an AI-powered analytics solution has to be able to scale to accommodate any workload. It must be able to ingest vast amounts of data quickly and prepare it for the predictive analytics tool. For example, OpenText™ Magellan™ uses a combination of industry-leading data management tools, Apache Spark and Hadoop, to achieve ingestion rates in excess of 100 times faster than Hadoop alone.

Scalability also applies to the number of people on the system. So your chosen solution has to be able to support a large number of concurrent users while ensuring the performance, security and availability of your automated analytics capabilities.

5. Usability

In most organizations, there will be three separate types of users requiring advanced analytics capabilities: Data scientists, knowledge workers and business users. Each has its own set of requirements for its data analytics tool and the best AI-powered analytics software contains features designed specifically for the individual groups.

  • Data scientists have access to analytics workbench functionality that allows them to train their machine learning algorithms and code their predictive models.
  • Knowledge workers can populate the predictive model with their own data, automate the running of the model and share the results with business users.
  • Business users have access to a wide range of reporting dashboards, charts and graphs that let them quickly and easily gain insight from the results.

Increasingly, advanced analytics solutions are offering more sophisticated self-service features. Self-service analytics allows business users to do most things themselves without the need for data scientists.

The best AI-powered analytics software will provide users with an analytics dashboard that allows them to query and manipulate large amounts of data and select how the results are presented. They can also share these results with colleagues, partners and other stakeholders.

6. Flexible deployment

According to Forbes, the next round of growth in the cloud will be driven by artificial intelligence and analytics. The ability to handle vast amounts of data and apply immense computing power to it is a strength of cloud-based services because you can instantly add more storage as needed, without adding more physical servers. This makes the cloud ideally suited for feeding data into the machine learning algorithms.

However, many organizations will continue to have security and compliance concerns about moving corporate data out from behind the firewall. The best AI-powered analytics tools allow for flexible deployment on-premise, in the cloud or a hybrid deployment where some advanced analytics tasks are conducted in the cloud and the more sensitive work remains in-house.

7. Open source

The concept of open source means the source code of a particular technology or solution is open for everyone to add to and improve. Open source machine learning, which Magellan leverages through popular technologies such as Apache Spark, Hadoop, and Jupyter Notebook, allows for communities of developers working together to address bugs and bring new feature sets and predictive models to the original solution quickly and cost-effectively. It allows for your AI-powered analytics software to benefit from the ecosystem of complementary open source solutions and developers.

Why choose OpenText for AI-powered analytics?

OpenText is at the forefront of practical applications for artificial intelligence software today with OpenText Magellan. This powerful data analytics tool uses AI in a business context to deliver improvements based on the data and systems you already have. It pulls together both structured and unstructured data throughout your organization and analyzes it to identify patterns.

Editor’s note: This is an installment in our “AI Glossary” series of blog posts, offering guidance on key areas of artificial intelligence and analytics. Look for future posts in this series over the months to come and read our first post in the series.


OpenText, The Information Company, enables organizations to gain insight through market-leading information management solutions, powered by OpenText Cloud Editions.

Related Posts

Back to top button