Although developments in machine learning are overwhelming the company, many challenges prevent teams from seeing the full value of these projects, Algorithmia discovered.
How technology improves the human side of customer service
Sales Patt’s Bill Patterson explains the three areas of focus of the company around speech technology for conversations with customers or employees.
More about artificial intelligence
An Algorithmia report published on Thursday revealed the challenges associated with increased use of machine learning in 2020. Most companies will be in the early stages of machine learning in 2020, but to move into more advanced stages come, organizations must overcome a number of report obstacles found.
SEE: The impact of machine learning on IT and your career (free PDF) (TechRepublic)
Algorithmia’s 2020 State of Enterprise Machine Learning report interviewed 745 technical professionals to determine how organizations plan to deploy machine learning in 2020, and the main issues associated with the journey. The biggest challenges associated with implementations of machine learning were scaling, version management, and budgeting, according to the report.
“AI and machine learning will be the most impactful technological progress that we will see in our lives,” said Diego Oppenheimer, CEO of Algorithmia. To assist organizations in their efforts in the field of machine learning, the report split their data into the following seven key findings:
Seven important findings
1. Emergence of data science for machine learning
“The role of data science is to collect a lot of data that these companies have collected and make sense,” and technological advances have led companies to generate more data, resulting in the need for more data scientists, Oppenheimer said. .
This increase in demand will continue until 2020 as machine learning occurs more and more: nearly 60% of organizations will employ between one and 10 data scientists, the report found.
More than half of those organizations have set up at least one machine learning project, but these implementations are expected to double by the end of 2020, Gartner thought. With the expectation that machine learning projects will increase, the report discovered that the company will see new data science job titles, including machine learning engineer, machine learning developer, machine learning architect, data engineer, machine learning operations and artificial intelligence (AI) operations.
2. Saving costs has priority
The report also looked at what companies want to get out of machine learning. Across the board, the three main use cases include reducing operating costs (38%), generating customer insights and intelligence (37%) and improving the customer experience (34%), the report said.
“Machine learning has the ability to reduce errors in many cases, which can help a company make more money and save money,” Oppenheimer said. “As in jobs where a lot of data is entered or processed, where many people may be involved, where it is error prone and somewhat slow, machine learning can automate many of them and make them more precise. It frees those people who do basic data entry to higher level, for which people are better suited. ”
Although medium-sized to large companies are mainly focused on cost savings, the report showed that small companies are more interested in improving the customer experience.
Smaller companies try to retain customers and have a permanent business – a problem that larger companies may not have. When thinking about how to use machine learning, optimization is a huge use case, Oppenheimer said.
For example, when a customer comes on the line and is bumped from one customer service representative to another, it becomes “a frustrating experience for everyone, and that is actually very expensive for the organization,” Oppenheimer said. “A lot of work is now being done in terms of asking questions, the agent on the other hand is typing it into a Google search box. A lot of customer care information is coming relevant to them, that is all that data science is used to serve the customer faster These are all things that improve the customer experience, making the customer ultimately (loyal). ”
3. Overcrowding at early maturity levels and AI due to AI
Machine learning projects will still be at an early stage in organizations in 2020: 21% of companies said they would evaluate use cases, and 20% identified themselves as early-stage adopters in the production of machine learning, it turned out from the report.
Respondents said they would be in different phases by the end of 2020. About 23% said they would work with models in production, and 22% said they would develop models.
“You won’t imagine a company in the future that doesn’t use machine learning and data science to optimize their business,” said Oppenheimer. “The problem is that many teams go in without understanding what the final result should look like. The truth is that you have to understand what business optimization should look like.”
4. Long way to commitment
It takes companies a long time to implement machine learning. For only one model for machine learning, respondents said they could spend up to 90 days on implementation. Nearly 20% of companies said they last longer than 90 days, the report found.
The process can take a while because machine learning projects are so new that current data scientists may not be fully familiar with the process, which goes back to why new data title functions are coming up in 2020, Oppenheimer said.
The road to implementation is longer at larger companies, according to the report. The main reason is that the larger the organization, the more approvals and people are needed to oversee the project, Oppenheimer said.
5. Problems with scales
Scale models were mentioned by respondents in the report as the biggest challenge (43%), an increase of 30% last year. This challenge is probably related to decentralized organizational structures, which according to the report often result in friction of tooling, framework and programming language during scaling.
“A big obstacle is that there is a lot of tools,” said Oppenheimer. “The employees who build the models are usually not the best people to use the scale. Organizations must (must) realize that these teams must have different skills and then be able to realize that the frameworks are moving very fast. The machine teaches space moves at lightning speed. ”
One solution that was offered in the report was the creation of innovation hubs within organizations. These hubs are dedicated to innovation projects such as machine learning and, according to the report, can work in an agile way to standardize the efforts of machine learning.
6. Difference between budget and maturity of machine learning
The budgets for machine learning generally increase, but vary based on the maturity phase of the project, the report found.
Companies that were halfway through the machine learning cycle increased their machine learning budgets by 1% to 25%, and 39% of companies in advanced stages of innovation did the same. About 30% of organizations with advanced levels of machine learning said they have increased their budgets by 26% to 50%, the report said.
What this data demonstrates is that “If you can prove success, you get a bigger budget,” Oppenheimer said. “We have seen companies across the board increase their budget for machine learning and data science. But for companies that do this the longest, they reach a certain maturity level where they need it in larger parts of their and therefore larger budgets to meet it to link. “
7. Determining the success of machine learning throughout the organization
The two most important statistics for determining the success of machine learning are business statistics and a technical evaluation of the performance of the machine learning model, the report found.
“Ultimately it’s about results,” said Oppenheimer. “Much of it is about involving people in the process of understanding the end.”
Teams must determine why they want to implement machine learning projects and find those end goals, rather than just implementing AI for AI’s sake, he added.
For more information, view AI and ML in the company on ZDNet.
Next Big Thing newsletter
Be up to date with smart cities, AI, Internet of Things, VR, autonomous driving, drones, robotics and more of the coolest technical innovations.
Delivered on Wednesday and Friday
Register today
