This can potentially expose sensitive employee information if not properly secured. Unauthorized access or data breaches could occur. Ans: Security is always a good concern. Employee and customer information is not used to train the OpenAI model. Also, we store separate indexes for each customer, so one customer’s information is never mixed with that of another customer. This data is maintained as part of our core data center, and subject to the same strict security regime as the rest of the data.
Ask HR provide incorrect or outdated information, leading to misinformed decisions or actions by employees. Ans: Ask HR always fetches information from the document you have trained it with. If you’ve trained it with the latest documents, the data will be up-to-date and correct. We are using the latest and most powerful model from OpenAI, GPT-4-O, which has background knowledge up to October 2023. That said, Gen AI is still a developing technology and there is always a probability that the response provided by AskHR will be incorrect or outdated. At this moment, that risk is considered in line with the business benefit that Gen AI confers, and most organizations find that the benefits far outweigh the risks. The risks are mitigated by an interface that promotes human oversight and follow ups with additional context.
Ask HR may struggle to understand complex queries or context, leading to frustration and inefficiency. Ans: Ask HR understands the semantic meaning of the user query very well. Also, it can understand non-English languages well. Therefore, it is unlikely that AskHR will frequently struggle with context or complex queries. On the contrary, Ask HR often outperforms most kinds of traditional searches, and that gap is only getting larger. In this regard, the capabilities of Gen AI represent a very significant leap from that of the time before the widespread adoption of Gen AI.
Over-reliance on the chatbot might reduce direct human interaction, which can be crucial for handling sensitive HR matters. Ans: Ask HR does not replace HR professionals but assists employees with their queries. This allows HR teams to focus on more complex and sensitive matters that require human intervention. Instead of replacing human interaction, Ask HR complements it by handling routine tasks and directing more complex issues to HR professionals, ensuring that human resources are utilized effectively.
Integrating the chatbot with existing systems may pose technical challenges and require significant resources. Ans: Ask HR is out of box integrated with the rest of the ClayHR, and the end customer does not require any specific technical capability. Regular HR professionals are able to use ClayHR out of box without any significant technical challenge. The underlying technical work is done by ClayHR’s own AI team, which consistently stays updated with the latest models and information to address technical challenges effectively. Further, our team is committed to supporting our customers in data integrations.
Compliance of the chatbot with all relevant labor laws and regulations can be complex and requires ongoing monitoring. Ans: We don’t recommend Chatbot / Ask HR to generate any labor laws and regulations or such compliance. Rather, we encourage its usage to provide a conversational layer on top of your already existing and already available regulatory and policy documents.
Employees might hesitate to use the chatbot, affecting its effectiveness and ROI. Will people be fully trained on chatbots? Ans: It's indeed possible that employees might initially hesitate to use Ask HR, which could impact its effectiveness and ROI. However, to address this, we provide detailed knowledge base articles and demo videos on how to use the Ask HR. This ensures that employees can easily understand and interact with Ask HR, building their confidence and encouraging adoption. Ultimately, this will help them in their daily operations and make addressing queries easier. But yes, this is a cultural issue, and people will usually follow each other’s lead on product adoption.
How often are updates and maintenance done to keep the chatbot functional and secure? Ans: We perform thorough testing of the chatbot to ensure that it is functional and secure. If any bugs or vulnerabilities are found in between, we take immediate action to address those issues as well. Additionally, if there are areas where the chatbot can be improved in terms of efficiency, accuracy, or security, we upgrade the chatbot with those as well.
