Exploring the use of generative AI advice for the academic advancement of faculty
This paper reports on the development and user testing of "AskADD," an AI-driven tool designed to streamline academic career guidance and enhance transparency in promotion and tenure processes for over 2,800 clinical faculty members at the SingHealth Duke-NUS Academic Medical Center.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
Imagine you are a doctor at a big university hospital. You want to know how to get promoted or earn tenure (a permanent job security). The problem is that the "rulebook" for this isn't a single, easy-to-read book. Instead, it's scattered across 30 different documents hidden deep inside the hospital's internal website. Finding the right rule feels like trying to find a specific needle in a massive, messy haystack, and it takes a long time.
To fix this, the researchers built a digital "smart assistant" named AskADD. Think of it as a super-fast, super-knowledgeable librarian who has read all 30 documents and can answer your questions instantly, 24/7.
Here is how they tested it and what they found, explained simply:
The Two-Step Test Drive
The team didn't just launch the tool; they tested it in two rounds, like a car manufacturer testing a new model.
The "Alpha" Test (The Prototype):
- They built the first version using standard, off-the-shelf tools (like a basic Lego set).
- They asked 35 faculty members to try it out.
- The Result: The librarian was helpful but a bit clumsy. Users often had to ask the same question twice or rephrase it because the assistant didn't quite "get" them. It saved some time, but it wasn't perfect. Users felt it was okay, but they weren't eager to tell their friends about it yet.
The "Beta" Test (The Polished Version):
- The team listened to the complaints. They upgraded the librarian to a "Custom AI" (a more advanced version of the brain behind the chatbot). They fixed the technical glitches that broke links to the original documents.
- They asked the same 35 people to try it again.
- The Result: The upgrade worked wonders.
- Speed: It got answers faster.
- Clarity: People stopped having to rephrase their questions.
- Trust: Users felt the assistant was more "empathetic" (it sounded more human and helpful).
- Recommendation: The number of people willing to recommend the tool to a colleague more than doubled.
What Did People Actually Like?
The researchers found that the faculty members loved the tool for three main reasons:
- Time-Saver: Instead of digging through the "haystack" of 30 documents, they got a direct answer in seconds.
- The "Map" Feature: In the second round, the assistant started giving direct links to the original rulebooks. This was huge because it let users verify the info themselves, which built trust.
- The "Human" Touch: The upgraded version felt less like a robot and more like a helpful guide, making people enjoy the conversation more.
What Was Still Missing?
Even though the tool was a success, the users had a few wishes for the future:
- Personalization: They wanted the assistant to look at their specific resume and say, "Here is exactly what you are missing to get promoted." The current tool only gives general advice because of privacy rules (it can't see private employee data yet).
- More Topics: They wanted it to cover not just promotion rules, but also general career development and HR questions.
The Bottom Line
The paper concludes that this "AskADD" tool is a successful experiment. It proved that you can build a secure, AI-powered guide that helps doctors navigate complex career rules without needing to hunt through endless files.
However, the researchers warn that this tool is like a specialized GPS built for one specific city (their hospital). It works great there because it was built on their specific maps and rules. Other hospitals might not be able to copy-paste it exactly because their "city maps" (policies) and "roads" (IT systems) are different.
In short: They built a smart, 24/7 career guide that saved time and reduced frustration. It's not perfect yet, but it's a massive step forward compared to the old way of searching through 30 different files.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.