As companies make contingency plans in case we find ourselves in a recession, they would do well to make sure they don’t sabotage long-term strategic projects such as intelligent automation by jettisoning the very people who are crucial to those efforts.
Rather, times of economic uncertainty demand companies invest in technologies that can deliver a reliable, ongoing return on investment, a topic I covered in a previous post. In order to do that, however, companies must resist the temptation to simply cut costs by reducing headcount without serious regard for the skills they let walk out the door.
In some instances, you may even find all the recession talk creates a good environment for investing in your people by training them in disciplines such as intelligent document processing that will be crucial to digital transformation efforts.
Related content: Data science and business groups need to speak the same language on intelligent automation
Crucial skills for intelligent automation projects
Let’s start with the most valuable performers on your automation Center of Excellence team. For my money, they are the people who work best with the business, who understand business processes and how to address them through automation projects.
Not everyone is capable of working well on big, transformational automation projects. Don’t sacrifice people who have been instrumental in helping you become a mature automation organization in the name of short-term cost-savings. And remember we are still in early days in terms of automation, so “maturity” may be measured in just a year or two. It’s important not to trim the folks who are helping you climb the maturity curve.
One such skill to keep an eye out for is the ability to map business processes and fine-tune them before automating (or even instead of automating, in some instances). By no means should any company be automating processes on a 1:1 basis between what employees are currently doing and the steps the automated process will follow. Rather, you need to identify where processes need improvement, or even wholesale re-architecting, then automate the new, improved version of it. That makes folks who are skilled in disciplines like process optimization quite valuable.
Other important roles are those having to do with oversight and governance. I liken this to cutting headcount from human resources, making it difficult for employees who may be having issues or looking to have simple questions addressed. Similarly, you need a governance function to continually ensure your automation models are working properly, even long after they’ve been launched. You don’t want an automated email responder sending erroneous or irrelevant messages to your customers, for example.
Automation skills to develop
Keep in mind many of your employees are likewise filled with trepidation about what a recession may mean for them. A good way to give them peace of mind, while also setting your company up for long-term success, is to offer training in skills like unstructured data automation.
It’s easy to find folks who are skilled at robotic process automation, which deals only with rather simple, structured data. Automating processes involving unstructured data is closer to the cutting edge for automation, making it more difficult to find people who are skilled at things like unstructured data extraction and writing business logic and rules engines.
Taking it a step further, offering employees education in intelligent automation skills like natural language processing will serve your company well.
Cloud is another hot area. Given a lot of automation tools run in the cloud, and/or use data stored in the cloud, upskilling employees on cloud computing and data storage will be to your advantage.
Related content: How intelligent document processing helps address the ambiguity in automation efforts
How to keep automation training costs down
And ‘lest you wonder where all the money for this training is going to come from, I give you my final point: upskilling your employees doesn’t have to cost a lot.
Lots of vendors offer training on their platforms that you can take advantage of. Indico Data, for example, offers training and certification on how to create workflows for unstructured documents. Customers generally have training hours built into their contracts, so the training is likely to be free or low-cost for most. Worst case, ask for a break; that’s exactly the sort of thing vendors are willing to offer up in the name of keeping customers happy during uncertain times.
You can also find lots of open-source tools for helping employees hone their skills with unstructured data automation. Platforms like Hugging Face and SpaCy offer guided explanations of what’s possible with language modeling, often for free.
Indico Data likewise has lots of documentation on how to build automation models in three languages: Python, C# and Java. Users who really want to get their hands dirty can visit our open source repositories, Finetune and Enso. Finetune is a library that allows users to leverage state-of-the-art pretrained NLP models for a variety of downstream tasks and Enso provides a standard interface for benchmarking transfer learning methods on natural language processing tasks.
Such tools provide resources that enable your employees to learn about important technologies such as NLP and how to automate unstructured data driven tasks – at the very least they’ll be more informed in conversations with vendors.