JengJengFin

The implementation of AI to help people that don't have the access to the bank loan system. We aim to be the top fintech to help the underprivileged.

Comment

Inspiration

Isn’t it all so overwhelming the loan application process whether having to gather all of the information about the assets, checking loan eligibility, calculating the monthly installments, and approaching various banks for the best deal. Without adequate tools to handle the complexity of this situation, it is normal to face inefficiencies and errors. Thus, we want to develop a database platform that can automatically calculate the debt system and effectively decrease the time spend for both banks and customers.

What it does

  • Generate loan agreements that protect both loanee and loaner.
    • Using KYC as a way to increase the protection of the users.
  • Use machine-learning and AI implementation to help valuation real-estate as the object to guarantee the contract.

How we built it

KYC section

  1. Face detection
    • The camera recognizes and tracks the image of a face, whether it is alone or in a crowd.
  2. Face analysis
    • The software reads the geometry of your face to identify the facial landmarks that are key to distinguishing your face.
  3. Data converting
    • Computer generates filters to transform face images into numerical expressions using deep learning to determine the similarity.
  4. Matching -The faceprint is compared against a database of other known faces. The determination is made when the results match others.

Real-estate valuation

  1. Create a model and gather data(Kaggle model)
  2. Train the model using the data.
  3. Input the data about the properties of the house them import the data into collab to run the code.

Challenges we ran into

  • Data can be hard to gather in real life because the data that we need for this model is private data. If the people don't want to give the data, the only way that we can do this is to start our service and gather the data from there which consumes a lot of time. So right now the only data that we could use is from an open-source such as Kaggle or GitHub which we cannot be sure that it is 100 percent accurate.

Accomplishments that we're proud of

  • Expand the horizon of our knowledge about the implementation of ai, machine learning, and blockchain technology.

What we learned

  • There are many outsource data on the internet that can be used to improve our startup or any other projects that we will experience in the near future.
  • There are still some aspects in that AI cannot be implemented to reduce bias.

What's next for JengJengFin

  • Improve our machine learning model from the data that can gather from our past customers.
  • Using this model to predict the customers' decisions in the future.
  • Analyse and evaluate the risk to barricade them before it happens.
  • Improve transparency.

Built With

Share this project:

Updates