ScholarlyAI

Intelligent Academic Workflow System.

Comment

Inspiration

I'm a PhD student specializing in cryptology, focusing on bridging the gap between connectivity and security for IoT (Internet of Things) and IIoT (Industrial Internet of Things). My supervisor's subject title was quite general, which made it challenging for me to find a clear starting point. There are many topics that intersect, including LWC NIST standards, PUF (Physically Unclonable Functions), Digital Twins, and Zero Trust Architecture for IoT, all of which relate to "bridging the gap between security and connectivity."

I needed a resource that I could access anytime I powered on my device. Typically, a human supervisor might hold meetings every 15 days, or ideally each week if they are particularly supportive. In my case, meetings every 15 days are not sufficient to address all my questions and provide the guidance I need. To solve this issue, I created ScholarAI—a supervisor that will never say no to you. It’s powered by Gemini Pro 3.

What it does

A smart platform that acts as both a supervisor or peer-reviewer to help research become much faster and more efficient. upgrading the performance of research without losing the human quality of researchers. (Giving smart mentoring, guidance, options ..) Using IA while not brainrotting a future researcher.

How we built it

Google Aistudio, Gemini-pro, coding.

Challenges we ran into

I am currently unable to access Google's paid features, such as Cloud, API, and Video Generation, during the testing period. This restriction limits my ideas to only what is available in Google AI Studio, making the competition potentially unfair. Nevertheless, I have enjoyed the performance of Google AI Studio, even while using the free version.

Accomplishments that we're proud of

  1. Coming up with an innovative idea. something doesn't exist yet in the market 1.1 Creating a Research journey map (Scispace (No), Elicit (No), Semantic AI (No)) (ScholarlyAI (Yes)) 1.2 Creating a Brainstorming to help a PhD supervisor pick up a relevant subject (Scispace (No), Elicit (No), Semantic Ai (No)) (ScholarlyAI (Yes)) 1.3 Creating a peer-review AI to shorten the time for both PhD students (waiting and getting a rejection) (wasting reviewers' time). So it's more practical to run the paper through a simulation to see how likely is it is to be accepted. Only better papers will tend to be uploaded via online journals everywhere. (Scispace (No), Elicit (No), Semantic Ai (No)) (ScholarlyAI (Yes)) 1.4 Creating a research space viewer to be able to see where researchers are congregated. (Scispace (No), Elicit (No), Semantic Ai (No)) (ScholarlyAI (Yes)) 1.5 Papers comparetor to be able to compare your paper with an expert's one to see whether it academicaly match or not to avoid rejection for things that are out of the research context, which is very common (grammar, design, etc) (Scispace (No), Elicit (No), Semantic Ai (No)) (ScholarlyAI (Yes)) Other features may be included but not mentioned.

  2. Creating a full functional but not ready to production application (Barrier of payment, can't integrate Stripe, PayPal for subscription,s etc.)

What we learned

  1. typescript to do some code editing (minor ones)
  2. prompt engineering
  3. using google-aistudio
  4. watching IA doing it's magic

What's next for ScholarlyAI

Having Google satisfied is essential for achieving success and securing funding, ultimately transforming our idea into a real project. I envision a team of experienced individuals who I will lead to accomplish the following goals:

  1. Create a centralized platform by Google, going beyond Google Scholar. This platform will feature unique IDs and web avatars for single identity management, catering to researchers worldwide. Those who prefer to reject centralization will be required to pay a commission for identity verification each time they wish to submit new research.

  2. Integrate all functionalities to enhance the research experience, making it more than just a scientific workspace, but rather a human-machine structured platform (IA).

  3. Google will earn a commission from all users who utilize premium features. The system will be automatically integrated into nearly all journals, meaning that journals will need to pay for auto-validation, thereby simplifying the lives of peer reviewers.

  4. Fraud will no longer be an option. Even if you personally know me, you will only be able to accept my paper based on its merit. Machines do not differentiate between people or races; they evaluate performance.

Built With

Share this project:

Updates

Submission history