11 hours ago
How digital labor as a service can benefit organizations and employees
Artificial intelligence and machine learning tools are becoming nearly ubiquitous in organizations across industries and geographic locations. As AI's capabilities mature and widespread adoption increases, businesses are discovering new strategies to improve their efficiency and productivity.
DIGITAL LABOR: THE NEXT WAVE OF AUTOMATION
'Companies are looking for ways to automate,' says Jay Bhatty, CEO and founder of a company that provides automation software to clients in the oil and gas industry. 'People often talk about how automation is affecting job security. Employees are nervous about losing their jobs, but there's actually a subsection of jobs that they do not like doing and they wish were automated away. These repetitive, mind-numbing, manual tasks can take up hours of employee time every week, and they are ideal candidates for automation.'
Converting nonstandardized data to standardized data, for example, can be a cumbersome and time-consuming process. Bhatty describes a scenario he's seen repeatedly with clients: A company's accounting department receives a 400-page invoice as a PDF—but before the team can process and pay the invoice, they must first convert it to an Excel spreadsheet with standardized formatting. In the past, employees needed to clean up this data manually or with optical character recognition (OCR) software.
'OCR scans every letter, one by one, but it doesn't understand the content of the document,' explains Bhatty. 'But now you can feed hundreds of documents into a large language model that is built to quickly read a massive document and understand its content. With an LLM, standardizing data in an invoice is very straightforward. It just adjusts a few items—date, line item, tax, total—and that's it. These types of tasks are ripe for automation, especially using LLMs, and employees will happily give them away so they can focus on more mentally challenging, revenue-generating tasks.'
began offering digital labor as a service to clients in 2023. In approximately 18 months, Bhatty and his team have built more than 1,000 bots for companies in the oil and gas industry to automate over 2 million hours of human labor. Bots are customized to perform specific tasks and can run concurrently with other bots executing their own independent functions.
'Clients like digital labor because it's a subscription-based service,' says Bhatty. 'It's fast, deployed in seven to 10 days, with a low initial investment. Tomorrow, if they don't need the bot, they can just cancel it or tweak it to do a different task. They don't have to invest millions of dollars up front in a huge monolithic system and hope that it works a year down the road. And because many of our clients are doing similar tasks, we can easily replicate processes to build new bots tailored to each company's needs.'
For leaders who are considering implementing digital labor as a service, Bhatty advises starting with these three steps.
Be aware of the sensitive information that bots and humans can access. Establish clear privacy and security parameters to protect against breaches.
'Bots can be programmed to deploy so that only certain individuals can see their data,' says Bhatty. 'If you're working with a software partner, make sure you know if anyone on their team can see your data. Our clients run bots in their cloud environment and control all their data—even we don't have access to it.'
2. GO FOR LOW-HANGING FRUIT
Bots are ideally suited to low-level automation. Identify straightforward tasks that bots can execute without critical thinking or decision-making.
'There are three main categories of labor: on-site employee labor, outsourced off-site labor, and digital labor,' says Bhatty. 'In this third category, a bot can run anywhere, anytime, without a person actively running the software. AI that can handle more complex decision-making tasks is on the horizon, but it will take time to develop. It's like trusting the technology behind a driverless car. For now, keep it simple.'
3. KEEP PEOPLE IN THE LOOP
Digital labor isn't intended to operate independently from humans. You still need people involved in conceptualizing, building, and fine-tuning bots, as well as resolving any problems that arise. Don't make decisions in a vacuum.
'A lot of employees and leaders are anxious about automation leading to job losses,' says Bhatty. 'The biggest eye-opener in launching digital labor as a service was finding the opposite, that a lot of people want a part of their job to be automated. Get input from your employees on the tasks that they dislike and take up a significant amount of their time. Look for opportunities to hand these responsibilities over to bots so your people can focus on more challenging and rewarding jobs.'
