Joining a new organization and getting to know the scheme of things is always a challenge, especially in a new field, especially if it is your first serious work experience. I think anyone who started any job can relate to this. However, if we talk about a complex, formal environment with multiple departments, each with their own processes, policies, and twisted interdependencies, the brain of a newly graduated intern will enter an information overload usually somewhere between the first hour and the first day of onboarding. This is undoubtedly a very widespread problem. What is more, a recent McKinsey report suggests that workers spend a quarter of their time searching for information.
So, what do companies actually offer to help you navigate all of this?
In my case, the problem was finding information about the different processes and practices in the company. As we live in an age of digital technologies, relying on digital resources seems more efficient than bothering colleagues. Therefore, I will describe my personal experience of dealing with the digital onboarding tools and instruments.
We had an internal system that worked much like Google, but exclusively for internal use. The issue was that while most of the needed information existed somewhere on that platform, the search function didn’t work quite well. Often, I had to go through dozens of documents just to find a tiny but important detail. This process was both time-consuming and inefficient. That’s why I was extremely happy when I discovered that we also had an internal chatbot.
As I worked in a financial institution, using personal AI tools such as ChatGPT was not an option due to strict security regulations and the risk of data leaks. Moreover, the organization didn’t have a license for any major AI systems, except Microsoft Copilot, which was only available with limited functionality. Therefore, the company had to develop its own internal AI chatbot, which is small-scale but well-protected.
I had great expectations. Perhaps that was the problem. I assumed that if a company creates an internal chatbot, it should be trained on all internal documents and have access to the same datasets available to employees. Additionally, it could be specifically fine-tuned to understand organizational policies, procedures, and regulations. You can imagine my surprise and later frustration when I discovered that the chatbot had no access to internal files at all. What was even worse, employees were even warned not to upload any internal documents into it, which to me sounded completely counterintuitive.
The questions I asked were not complex: they were simple, practical ones any new intern might have: about internal systems, processes, or regulatory frameworks. I did receive general answers similar to external LLMs yet, the chatbot could not use the context of the organization appropriately because of its limited access. I understand that organizations have to be careful with data security, and perhaps developing their own LLM is cheaper or safer than licensing an external one. I would also not risk giving a tool like ChatGPT direct access to confidential databases. Still, if you already have an internal AI system, why not take advantage of one of the things that makes it superior to public ones? Why not give it safe and protected access to internal knowledge? Not using this capability felt like a missed opportunity. Moreover, it slowed down my integration in the organization as a newcomer.
I would genuinely love to hear from others who have experienced similar situations. Has your organization developed its own chatbot or integrated an existing AI solution into its infrastructure? And if so, how well does it actually help employees find the information they need?
