Mehardeep Singh, Shriyans Ghosh, Yueming Li and Mengzhen Xue, Zoya Hammad,

Emerald Honor for GENESIS I

Emerald Honor for GENESIS I

Emerald Honor for GENESIS I

International

-

Completed in

2023

Here are our Emerald Winners for GENESIS I! We received so many outstanding projects and the judges wanted to give an honorable mention to a few of them!

Note: The Emerald Winners are not in any particular order.


“Would you like to buy fruit?”

As the project title suggests, you would think that Yueming Li and Mengzhen Xue have created an aesthetic application for produce lovers to shop for their favorite fruits. But in reality, they have created an intricate and strategic platform for victims of domestic violence and abuse to save themselves. 

Under the guise of a fruit shopping website, victims of domestic abuse are able to access a secure site to access lifesaving information and get into contact with law enforcement without alerting their abuser. Through features such as a shopping cart, food surveys, and consumption logs, each serve a purpose in allowing victims to provide valuable information about their status. This is coupled with a uniquely classic UI/UX design which makes this tool even more inconspicuous and discrete for victims of violence. 

The judges at EQ Hacks were extraordinarily surprised to see a project guided by a clever strategy in approaching the issues of domestic violence. Although technical complexity is important, the project mission is equally impactful as well.  And we urge future participants to follow Li and Xue’s example: “[learning] about the importance of user-centered design in sensitive situations” 

For the future, the EQ Hacks team is excited to see the team develop a secure database system to protect client data—especially for the most-likely vulnerable users of “Would you like to buy fruit?”

Check out Yueming Li and Mengzhen Xue's project here:

  1. https://devpost.com/software/would-you-like-to-buy-fruit


Aicura

Diabetic retinopathy, otherwise known as (DR) is a serious eye condition caused by damage to the retinal blood  vessels due to prolonged high blood sugar levels in diabetics. This condition can lead to vision impairment and even blindness if left untreated. DR presents a significant health problem because it generally progresses without alarming early symptoms. And according to the NIH, around 103.12 million adults worldwide suffer from DR.  That’s where aicura, created by Mehardeep Singh, comes in.

Aicura utilizes the power of image processing algorithms and computer vision in order to efficiently analyze medical imaging data. By training a powerful ML model, aicura can help detect DR and other kidney diseases swiftly. At EQ Hacks, we found that this will undoubtedly impact the ability for doctors to reach an early medical diagnosis and create an effective treatment plan. 

Similar to the many other developers in the AI + Health space, Singh needed to create a project that minimized the chance of producing a false diagnosis. He had to ensure that variations within medical imaging and the manifestation of DR itself was accounted for by Aicura. Moreover, Singh wanted the ML model to be able to interpret different modalities: MRI, CT, etc…

The EQ Hacks judges were also thoroughly impressed by Singh’s interdisciplinary approach of combining ophthalmology and medicine with technology. His specific focus on DR and kidney conditions made it a more compelling and intricate project. 

Thus,  we at EQ Hacks are very excited to see this technology expanded into covering more severe, yet hard-to-catch medical conditions. Because as Singh says, we must all  learn to “[empower] healthcare through AI innovation.”

Check out Mehardeep Singh's work here:

  1. https://devpost.com/software/aicura


Quizicate AI

Every year around May, millions of high school students all across the US and around the world cram for  CollegeBoard’s AP Exams. For Shriyans Ghosh, studying for such a high-stakes exam in a limited amount of time can be draining. To make the review process just a bit smoother for him, Ghosh created Quizicate AI, an AI tool that takes  uploaded course material and customizes practice resources out of them. 

If a student were to use Ghosh’s tool, they’d be able to generate summaries, quizzes, and a key concepts document that is contained in any of the course content. However, beyond this innovative feature our judges were most impressed with the infrastructure that Quizicate AI was developed with. Ghosh utilized the AWS Amplify framework to construct the webpage. On the backend, he used CloudFront and Cognito to implement authentication systems and actually generate review material. Furthermore, in an attempt to prevent AI ‘hallucination,’ Retrieval Augmented Search (RAG) was used. 

The applicability of Quizicate AI extends beyond studying for AP tests or being limited to creating practice problems. That’s why Ghosh is expecting to allow Quizicate AI to create flashcards as well as short form reviews as well.

Check out Shriyans Ghosh's project here:

  1. https://devpost.com/software/quizicate-ai-i9yuma


EcoHabitat

From all the way across the world, we have Zoya Hammad of the Ivory Coast. Hammad’s submission to Genesis I was EcoHabitat, an educational platform that helps individuals practice sustainable energy use with common household appliances.

According to the United Nations, energy consumption will soon eclipse current production and will contribute to more and more carbon emissions. This is why the UN has established three Sustainable Development Goals addressing it: #11) Sustainable Cities and Communities; #12) Responsible Consumption and Production; Climate Action. Hammad’s project tackles these three emerging challenges head-on.

Hammad proposed 4 clear-cut features as part of EcoHabitat. Through ML5, an image classifier was implemented. This image classifier takes a reference image of a household appliance and outputs the estimated energy consumption per hour. This data includes not only maximum and minimum energy use, but also how much electricity is used when the household appliance is on standby. The EQ Hacks judges were particularly interested in this feature because it provides an efficient way for users to access information about their own energy use. Additionally, EcoHabitat also contains a form-based WattCalculator for users to self-track their energy consumption, as well as a user profile system. The user profile system also utilizes a “streaks” system to encourage app use.

Although EcoHabitat is a great start, Hammad acknowledges the improvements necessary for the image classifier. That’s why she hopes to train her own model for the WattSearch feature. Additionally, she is excited to improve the dataset that EcoHabitat is based off of by including more household appliances used around the world.

The EQ Hacks team sees EcoHabitat as a unique and surefire way of reaching and even exceeding the UN-set SDGs—something crucially needed in this time of climate change and growing energy usage among the world’s population.

Check out Zoya Hammad's project here:

  1. https://devpost.com/software/ecohabitat-8u2iej


SPIF

Our final Emerald Project is SPIF by Adrian Pham and another student*, which combines Artificial Intelligence and Clinical Pathology to help diagnose multiple myeloma (MM) in patients. 

Multiple myeloma is a type of blood cancer that originates in plasma cells, a kind of white blood cell responsible for producing antibodies. This malignancy leads to the accumulation of abnormal plasma cells in the bone marrow. This accumulation  will then  interfere with the production of normal blood cells and result in bone damage, organ dysfunction, and immune system impairment. According to the American Cancer Society estimates, about 35,780 new cases of MM will be diagnosed in 2024. Like most cancers, early detection is the key to survival and that’s where we see SPIF’s usefulness.

SPIF analyzes patient data from an array of medical tests in order to accurately diagnose a patient with multiple myeloma. Specifically, the SPIF conducted research and identified serum protein electrophoresis as one of the diagnosis methods. The SPIF AI detector also picks up on immunofixation data as part of its diagnosis algorithm. These highly specialized algorithms were built through Python and a variety of imported libraries. Ultimately, by increasing accuracy and efficiency, SPIF reduces expenditure on diagnosis by around 8 million dollars yearly.

The EQ Hacks judges were  most impressed by the supplemental research done by Team SPIF about multiple myeloma. Although AI and ML-based projects are becoming much more popular, Team SPIF integrated their knowledge of AI with clinical-based algorithms for cancer detection. 

*We were unable to find the registration data linked to the second participant’s name. If you are from the SPIF team, please email us at support@eqhacks.org with your official name to update your article.

Check out the SPIF project here:

  1. https://devpost.com/software/spif-a-novel-software-for-diagnosing-multiple-myeloma


If you liked this article and enjoyed reading about our honorable mentions, feel free to leave a comment of support on their DevPost page.

If you want the chance to get your own featured article, join our ongoing competitions.

Perfect for the visionary and realists.

Drop us a line with any questions, inquiries or business proposals.


we love helping.

Perfect for the visionary and realists.

Drop us a line with any questions, inquiries or business proposals.


we love helping.

Perfect for the visionary and realists.

Drop us a line with any questions, inquiries or business proposals.


we love helping.