Exploring New Uses of Chatbots for Child Welfare (Volume 4)
As chatbots based on LLMs (Large Language Models) develop and improve continuously, the possibilities for how they can help improve the work of child welfare professionals expands. See a few new examples below of how various roles within a child welfare agency get use these tools to achieve improvements in efficiency as well as the strategies we employ for doing our jobs.
- Iterating on Case Plans
Chatbots can help us come up with ideas we may not have thought of previously. They can be our “thought partners” by helping us come up with ideas and navigate through difficult situations with complex needs. They can highlight different solutions that we may not have properly considered.
What to do: Describe a specific case or situation in your prompt (always making sure to remove any names, dates or any other information that could expose identities). Start with a basic description and then iteratively add details each time to see how the chatbot’s response improves. Ask the chatbot what other critical information you might be leaving out so it can help you think through possible courses of action. You can even ask it to debate (with itself) about the pros and cons of different strategies that it has suggested.
- Evaluating Report Quality and Clarity
Fixing errors or inconsistencies in reports of any kind can delay the processes that help us provide care to children and families. Using chatbots to evaluate your work before you send it to its recipient(s) can save time and act as a proofreader or reviewer.
What to do: Upload or copy in your report to the chatbot and ask it to provide editing suggestions that would make your report more clear. Ask it to take evaluate your work as if it were a supervisor in your department, a family needing to understand information, or a judge wanting to understand a particular case. If there is flexibility in how you present the information, ask the chatbot to provide suggestions for improving the format overall (then ask it to produce this format).
- Process Improvements through Reverse Engineering
Drawing on oodles of data points from across the internet, chatbots can provide the expertise of the best process engineers to any situation if you give it the right context. Even expert process engineers can benefit from vetting and exploring their ideas using a chatbot before they implement them.
What to do: Start by describing the outcome your department wishes to achieve and ask the chatbot what strategies and tactics may have led to this improvement. As it lists these ideas, ask it to describe particular ideas in more detail. You might suggest, for example: “Family recruitment staff increased foster home recruitment by 20%. What strategies and tactics might have led to this improvement? Please include software tools as well as general workflow improvements that may have contributed to this outcome.” You can also ask it to evaluate your current processes to help find errors or inefficiencies that may have been previously overlooked.
These are just a few examples of some more ways that chatbots can help child welfare professionals do their jobs more effectively. As these tools continue to develop there will be even more beneficial applications to all sorts of situations that arise within child welfare. The best way to discover these is to get started with using them and explore them a bit further each day until they become part of the way we do our work.