Technology is evolving at a rapid clip. Staffing firms must scramble to stay on top as artificial intelligence, online staffing, chatbots and other automation evolve. New online platforms are “hot;” nascent technologies are emerging daily.

While technology will likely never upend the staffing industry as the automobile did the horse and buggy, it is — to say the least — bringing changes.

For example, AI has been a buzzword for a while, but its potential includes use of machine learning and natural language processing to examine vast amounts of data in order to find workers or better match them with jobs. AI is able to take “unstructured data” to find people who have the right skills but whose résumés don’t match the key words of a job description.

“What impacts me the most and keeps me up at night right now are recruiting challenges and talent shortages,” says Jason Leverant, president and COO of AtWork Group based in Knoxville, Tenn. And AI is a tool that can help meet recruiting challenges.

Even while AI is good at prequalifications and sourcing, Leverant cautions that humans are still needed to determine if someone is the right fit for a job. Instead, AI lets teams work more efficiently and effectively by taking the repetitive and replicable tasks off humans’ plates.

“This just adds another arrow in their quiver,” he says.

AI isn’t the only technology trend: There are chatbots to communicate with workers or buyers. Further, some firms are giving Amazon’s Alexa a try as a communication tool. Still more have mentioned seeing changes at the VMS level. And then there’s what some consider one of the biggest trends affecting the industry: online staffing.

While not a scientific study, here are some of the top technology trends Staffing Industry Analysts sees and what they mean for staffing:

AI is Not Replacing HR; Not Now, Not Ever

An aim of AI is to automate recruitment to let HR focus on other things such as onboarding and employee engagement, says Carisa Miklusak, CEO of tilr, a mobile-first job-matching platform in Cincinnati with a development office in Toronto.

“Today, recruiters will tell you they spend hours going through résumés from job boards,” Miklusak says. And even when allowing an applicant tracking system to take a first look at the résumés, there is still a cost to the process as well as a lot of labor that goes in.

AI can automate that process to a large degree and allow firms to be more informed, with more data, to make better decisions, she says.

“I don’t believe that AI’s intention is to replace recruiters,” Miklusak says. “What AI is intended to do is to provide more data and more exact results.”

Tilr aims to match workers to job opportunities based on skills and not titles. Its algorithm is gender and name agnostic. Each time tilr sends out a worker, there is a feedback loop that comes from a human on how the worker performed, and that adds to the system’s knowledge about the appropriate workers to send in the future.

Workers sign up for tilr by downloading either an IOS or Android app then inputting their own skills. Once submitted, the candidate undergoes a background check, then an interview with a live tilr specialist. Next year, the company plans to use chatbots for interviews.

Client companies then order workers based on roles, such as a picker/packer at a warehouse, and number of workers needed. Then tilr sends the workers without the need to interview or select the workers ahead of time themselves.

The company is now focused on onsite 1099 workers in segments such as logistics, retail, warehousing and customer services.

In addition, the company is working toward true AI.

“What that means is that we’ve built an algorithm that makes decisions on how to staff roles based on factors we’ve taught it as humans,” Miklusak says. “As it makes decisions using this data, we have a 360-degree feedback that tells us if the decisions were good or bad, and the relationships that drive these decisions are strengthened or weakened, and the algorithm begins to learn. Right now, we are purposely controlling when we ‘teach’ the algorithm to make decisions differently and then automating to measure effectiveness. As we gather more data and the algorithm becomes smarter, we will allow it to learn on its own and alter its behavior as a result.”

In the Cloud: Growing Awareness

Staffing Industry Analysts estimates that human cloud firms processed spend of between $47 billion and $51 billion in 2016.

Human cloud includes firms using online platforms to bring workers together with clients. And they come in three flavors: online staffing firms such as Upwork; online work services firms such as Uber; and crowdsourcing businesses such as Amazon’s Mechanical Turk.

Workers often note the flexibility of online jobs as a plus.

Awareness of online staffing and the human cloud has been growing in the staffing industry. A survey by SIA found 34% of staffing firms did not know about online staffing in 2012. But a similar survey this year found only 4% were unaware.

SIA’s latest survey even found 14% of staffing firms currently own or have invested in a human cloud firm. One example: Adecco recently rolled out its hospitality and event online staffing platform Adia in the UK. On the flip side, 52% of staffing firms said they know of human cloud firms but had no interest in taking part in the segment.

Online staffing is one of the largest technology trends affecting the staffing industry, says Robert Hankin, senior VP of IT at Roth Staffing Companies based in Orange, Calif. Although he encourages the industry to embrace this trend, it does concern him that some online staffing services try to remove the human element of the staffing relationship.

Self-serve. “It’s like an automated self-serve staffing service representative mixed with a bit of VMS-type management capabilities from the customers’ perspective; mixed with job boards from the candidates’ perspective,” Hankin says. But in staffing, technology is simply there to help, he continues. “If you remove the relationship, then the technology becomes irrelevant. People come to staffing, whether it’s online or in person, for the human element … to have a partner who can help walk you through the process and be your advocate.”

If done right, online staffing can offer a lot of benefits, he notes. For example, it allows credit card prepayment for services, which is good for smaller, mom-and-pop organizations that might struggle to get credit otherwise.

Some of the technology now used by online staffing firms will likely migrate to traditional firms. Hankin says traditional staffing may take more of a hybrid approach, but it will be an evolution, not a revolution.

Inflection point? Is the human cloud at an inflection point? SIA President Barry Asin said the jury is still out during his keynote speech at the Collaboration in the Gig Economy conference last September in Dallas.

Online staffing has been out for about 10 years, and SIA research found 13% of large North American firms that use traditional staffing now also use online staffing. That’s about the same percentage of large buyers that used VMS 10 years after it came out. Currently, 71% of large buyers now use a VMS. However, whether online staffing use ramps up to the same extent is to be determined.

But staffing firms need to think about how to respond to online staffing and the human cloud, Stanford University Business Professor Paul Oyer said in his keynote address at the conference.

Oyer said executives at traditional staffing firms must ask themselves what they can do that platforms cannot and customers will pay for. Traditional firms must move up the “food chain” and add value — doing more than just introducing workers.

In an example of moving up the food chain in another industry, Oyer cited a ticket reseller whose business was disrupted by StubHub.com. To survive, the business moved from just reselling tickets to helping clients by planning and organizing entire weekends around events.

Tech Can Matter When Selling a Staffing Firm

Does staying up to date with technology matter when selling a staffing firm? Yes.

“The buyer will certainly want the seller to have up-to-date technologies and software in place,” says Akash Taneja, managing director of investment banking firm De Bellas & Co. “It just shows that they are not living in the Stone Age, the numbers they are providing are accurate and the accounting they are doing is being done properly.”

Buyers like to see sellers make investments in the growth of their company — technology being one of those areas that lead to growth, Taneja says. Also, it’s not just accounting but also applicant tracking systems.

In addition to providing confidence to the buyer, it can help with strategic buyers who may want to integrate the company into its back-office system.

Taneja also says some staffing firms have built their own proprietary software or bought software that they have customized for themselves, and that can be looked upon favorably by buyers because it is a unique property of the firm.

Does having an online staffing strategy make a difference with potential acquirers?

“The short answer is that we have not seen that play a role in the transactions that we’ve worked on,” he says.

AI Hype

Some of the top functionalities for artificial intelligence include finding and matching candidates, including through large amounts of “unstructured data,” such as résumés.

“While this concept may sound relatively simple, the core problem is that the average job posting and the average résumé tell two different stories,” according to the report, “Artificial Intelligence in Talent Acquisitions,” produced by SIA in collaboration with Talent Tech Labs. “One tells about the company, exploring what skills and competencies matter most in a specific occupation. The other tells a story of career history and accomplishments from an individual perspective. By analyzing the résumé and pulling out relevant pieces of data, the system can match up the skills with the job requirements to see if the individual candidate is a match in ways that are far more insightful and accurate than finding word matches.”

One caveat: The report notes that it’s easier to find examples of AI in hype than practice and some current AI applications are reliant on human intervention.

“At base, for a system to exhibit artificial intelligence, it should be able to learn in some manner and then take actions based on that learning,” Omar Abdelwahed, SoftBank Robotics America’s head of studio, was quoted as saying in an article on Futurism.com.

Sourcing, Then Engaging

“The world of sourcing has really been heavily augmented by the use of technologies,” says Brian Delle Donne, president of New York-based Talent Tech Labs. However, “when you generate a lot of candidates like that, the next challenge is trying to engage them.”

Right now, we’re seeing only emerging uses of machine learning to engage job candidates, Delle Donne says.

An example of engagement: Machine learning can be used in candidate relationship management software to monitor jobs websites to learn what content engages visitors and serve up more to keep them engaged.

But Delle Donne says that while AI (of which machine learning is a part) is here to stay, matching and prediction will be their main uses for the time being rather than engagement.

There is also candidate management software that can help staffing firms engage workers, but there are only a few products out now, he says. Some staffing firms use their ATS as a CRM tool, but they are not ideally suited for the function.

Engagement is especially important if you are putting together a talent cloud, Delle Donne says. “Unless you engage people, you don’t have a talent cloud. You don’t know if people are available, are a good fit, or will take your call.”

Chatbots for Employees

A chatbot, which can be powered by AI, is a computer program that can simulate human conversation. An example of a chatbot is Wade & Wendy, which received an investment from the Randstad Innovation Fund. The chatbot works with both candidates and buyers. In another example, The Adecco

Group partnered with Mya Systems, creators of Mya, an AI-enabled chatbot that aims to automate outreach, screening and communications with job candidates and interfaces with the company’s ATS.

Use of chatbots for interacting with employees is fairly new, says Ahmar Abbas, VP of global services at MSP and IT staffing provider Digital Intelligence Systems, or DISYS, based in McLean, Va.

“A lot of retailers have had them, but they are very basic,” says Abbas, who is rolling out a chatbot for DISYS that will communicate with workers.

Similar to chatbots are voice-based assistants such as Alexa on Amazon’s Echo. Abbas said his team is working on this as well and workers can already use Alexa to find the nearest DISYS office and there are plans to enable workers to search for jobs through Alexa.

More Mobility

Abbas sees other tech trends as well, including the increased use of mobile apps that enable workers to submit timesheets and find contact information. And they will evolve to provide even more functionality.

On the internal side, a trend he sees are MSP and VMS systems allowing more access to application programming interfaces, known as APIs. This lets staffing firms develop connections to the MSP or VMS electronically rather than having to cut and paste data from one system to another.

What to Watch

While they may not be trends in staffing yet, robots do have the potential to affect the industry. Pepper the robot’s appearance at SIA’s Executive Forum in Europe was very well received. But automation can affect the workplace, in other ways as well. Deaths caused by robots — 61 in the US workplace between 1992 and 2015 — has resulted in a new center set up to study their safety implications.

Another potential future trend is blockchain, the technology behind cryptocurrencies such as Bitcoin. Fortune describes blockchain as a way to create tamper-proof records across a large network of computers. The idea is that with records on so many computers, it would be difficult for one entity to be able to hack all the records. Several experts say blockchain is one of the up-and-coming technologies to watch.

However, Dhar Patadia, CIO of IT staffing firm Collabera, based in Basking Ridge, NJ, says blockchain, and AI as well, are still in the hype curve.

“When it gets into the mainstream we will see the impact on staffing,” Patadia notes.

Still, he agrees that AI, fueled by big data, will help streamline the flow of candidates to the right recruiters. However, “the human element is always needed because it’s a people business,” Patadia says. “To see that go away is going to take quite a bit of time.” But people will eventually get more used to robotics and automation such as chatbots; people have gotten used to other systems such as the phone systems used by banks.

Ultimately It’s Us

Several executives interviewed for this article stress that while technology is an important tool, it’s the humans that matter.

“To me, we’re in the people business; people still need to be involved,” says Hankin of Roth Staffing.

Even Google’s AI, considered to be the most advanced, is rated in the intelligence category of a 6-year-old. And while AI will become more advanced and eventually be able to play a very helpful role in the pre-screening/pre-hire process, it will take a while.

In addition, one of the old-fashioned methods of finding job candidates remains on top.

“Word of mouth is still the No. 1 way,” Hankin says. “People need to use those fundamental skills from talking to find those candidates; that’s never changed.”