If you are involved in HR tech you probably have enough on your plate without digging into some brand new technology.
And yet… important new tools are continually being released that you probably ought to understand.
One of the hottest new tools is Microsoft’s AutoGen, an automation tool leveraging the power of large language models.
But does HR ‘really’ need to know about it?
Let’s see if we need to look more closely at this tool.
What HR needs to know about AutoGen?
Before we get into any details about AutoGen, let’s step back and consider what you need to know:
- HR leaders need to have a sense of what it is capable of, why it is different from our existing tools, and how much effort it would take to start using it.
- HR professionals need to decide if it will help make automating processes much easier or if it can do automation that was not possible in the past.
Given all the other automation, tools why bother with AutoGen?
The reason Autogen is of interest is that it’s a forerunner of a new class of automation tools that will change how programming is done across the organization.
It makes things possible that were not possible before and does things in a way that is unfamiliar, even uncomfortable, for those with any familiarity with traditional programming.
Take a breath and read what Microsoft’s website says about it: “AutoGen provides a multi-agent conversation framework as a high-level abstraction. It is an open-source library for enabling next-generation LLM applications with multi-agent collaborations, teachability, and personalization. With this framework, users can build LLM workflows.”
If that’s not clear to you then that’s just a signal of how fast things are moving.
We are in an age of technology where it is hard to understand what’s meant to be a simple explanation of the tool.
That’s why we need to start paying attention.
About AutoGen
There is a lot AutoGen can do but it’s easiest to think of it as a way of building programs or automations using natural language rather than code.
This means that in theory, anyone can use it for coding or automation.
In practice, it’s more that someone can learn to use AutoGen much faster than they could learn a programming language like Python.
Furthermore, once someone is familiar with AutoGen, they will, in many cases, be able to create applications faster with AutoGen than they could by writing code.
My go-to AI advisor, Fuad Miah of Uncanny Labs, says that with this kind of tool, a project that might have taken months can be completed in a few days.
Just to get concrete for a moment, here are the sorts of applications you might build:
- Continually monitor data on labor markets and inform you of changes relevant to your business
- Automate onboarding processes
- Monitor the organization’s career page to make sure it’s working as desired
The spooky side of AutoGen is that it has the intelligence of a Large Language Model built in.
This means you might say: “Report on anything problematic about the applications coming in through our career site” without the need to specify what “problematic” might mean.
Now, you are right to think “it won’t be quite that easy”, and that’s certainly true, but it does have intelligence and hence it can do things you wouldn’t even attempt with a traditional automation tool.
Even more spooky is that rather than being a single entity, you can use AutoGen to create several different agents, each pursuing their own goals and talking to each other as needed.
It will take some hands-on experimentation to make sense of this multi-agent capability.
The first steps to moving forward
If you are an HR professional with a programming background and free time this weekend you might just dive in and learn all you can about using AutoGen.
There are endless resources online and, with a few days of investment of your time, you can become HR’s AutoGen expert.
If you are an HR leader and none of your staff have volunteered to spend their weekends learning AutoGen, then you’ll have to find someone who can coach you specifically on where it might be helpful in your department.
Check first with IT, consider asking a technology consultant in for a day, or assign the task of learning AutoGen to one of your staff as part of their job (not a weekend hobby).
All you want out of this first step is to have a good idea if there are enough high-value applications that it makes sense to take it further.
Conclusion
We are entering the Wild West of programming and automation as AI-based tools come swooping into our lives.
It’s best to get up and take a look at this thing that is arriving, so at the very least you have a good sense of its size and power.
There will be a lot more automation in the future, you will need to be good at managing it. This means understanding these complex and unfamiliar new tools.
It will be both fun and daunting.
Good luck.