Simret.exe
Hello Everyone π This weeks project I will be working on such an interesting project that I couldn't imagine I will be working on this Fintech filed. I will share more on it when I start building ππ«‘
Hey guys , How was your day ?
I couldn't get enough time to share in detail what I've been working on lately, but now that I've finally submitted and completed the project, let me walk you through it.
Since it's a Machine Learning project, there isn't much fancy UI to look at π . The interesting part is the idea behind it.
Imagine services like Telebirr, mobile banking, digital loans, and online payments. Every day, thousands of transactions happen, and financial institutions need a way to identify customers who might be risky before offering loans or other financial services.
So I built a Credit Risk Prediction System.The system:
->Analyzes customer transaction history
->Learns patterns from historical data
->Estimates customer risk levels
->Provides predictions through an API
Think about it this way. When someone buys airtime, sends money, pays utility bills, uses Telebirr, or performs other digital transactions, all of those activities create a transaction history. The model studies these patterns and estimates how risky or trustworthy a customer is.
One thing that excites me is that this project was built using transaction data from another country. It makes me wonder what insights we could discover if we trained a similar model using actual Ethiopian transaction data from services like Telebirr and local financial institutions. π€
I couldn't get enough time to share in detail what I've been working on lately, but now that I've finally submitted and completed the project, let me walk you through it.
Since it's a Machine Learning project, there isn't much fancy UI to look at π . The interesting part is the idea behind it.
Imagine services like Telebirr, mobile banking, digital loans, and online payments. Every day, thousands of transactions happen, and financial institutions need a way to identify customers who might be risky before offering loans or other financial services.
So I built a Credit Risk Prediction System.The system:
->Analyzes customer transaction history
->Learns patterns from historical data
->Estimates customer risk levels
->Provides predictions through an API
Think about it this way. When someone buys airtime, sends money, pays utility bills, uses Telebirr, or performs other digital transactions, all of those activities create a transaction history. The model studies these patterns and estimates how risky or trustworthy a customer is.
One thing that excites me is that this project was built using transaction data from another country. It makes me wonder what insights we could discover if we trained a similar model using actual Ethiopian transaction data from services like Telebirr and local financial institutions. π€
π₯4π3
Forwarded from BΟ
ΙΎΤΞ±Ι³-Ops (BΟ
ΙΎΤΞ±Ι³ | Ψ¨Ψ±ΩΨ§Ω :))
Your work can be amazing and still get ignored if nobody sees it. :))
π₯1
Forwarded from Bytephilosopher
CURSOR was giving away 1 year of Pro FOR FREE with student email
but Ethiopia isn't Eligible π
@byte_philosopher
@byte_philosopher
π2π’2
Forwarded from ππππgΠ½t
β€3π1π«‘1
Yap Last week we said It is new month ππ let us say also this week in our calendar π
π
Good morning pplπ«‘
Don't wait to work on your goal to be a new week or new month consistency β¨οΈ is the key to all of it.And also don't forget to enjoy life π«‘π€βΊοΈβΊοΈ
Have a good day all of youπ€π€
Good morning pplπ«‘
Don't wait to work on your goal to be a new week or new month consistency β¨οΈ is the key to all of it.And also don't forget to enjoy life π«‘π€βΊοΈβΊοΈ
Have a good day all of youπ€π€
Forwarded from BΟ
ΙΎΤΞ±Ι³-Ops (BΟ
ΙΎΤΞ±Ι³ | Ψ¨Ψ±ΩΨ§Ω :))
stop searching for approval in others. you are the approval.
β€2
How many of us know the word fraud ? Fraud is when someone dihonestly use your identity to steal from you or from your account.
I am building a Machine Learning system that detects fraudulent transactions. If someone sign up for an account and immediately buys smt expensive at 3 am from different country the system catches it and blocks it before the money is gone.
What I have done so far cleaned and explored over 400,000 transaction records and also balanced the data using SMOTE to train the model properly.
The next step is training a Machine Learning models including XGBoost. Evaluating which model catches the most fruad .
I learned so many new things and I will learn so many things from this project at the end.
#MachineLearning #fraudDetection
I am building a Machine Learning system that detects fraudulent transactions. If someone sign up for an account and immediately buys smt expensive at 3 am from different country the system catches it and blocks it before the money is gone.
What I have done so far cleaned and explored over 400,000 transaction records and also balanced the data using SMOTE to train the model properly.
The next step is training a Machine Learning models including XGBoost. Evaluating which model catches the most fruad .
I learned so many new things and I will learn so many things from this project at the end.
#MachineLearning #fraudDetection
π2π2
Forwarded from Eyu Devs
seeing people like @selfmadecoder and @Dagmawi_Babi try to get in to YC , and the one who made it like @bekacru_c is motivating AF.
glad to see such kind of pioneers , and I'm pretty sure that i will try soon.
glad to see such kind of pioneers , and I'm pretty sure that i will try soon.
"The sky would be the limit if there weren't footsteps on the moon"
π2β€1
"You should never regret the things you do. You should only regret the things you don't do."
Word of the dayπ₯
Word of the dayπ₯
π₯2
Simret.exe pinned Β«How many of us know the word fraud ? Fraud is when someone dihonestly use your identity to steal from you or from your account. I am building a Machine Learning system that detects fraudulent transactions. If someone sign up for an account and immediatelyβ¦Β»
Forwarded from Apleo Scholarships Ethiopia πͺπΉ
π Build with Gemini XPRIZE Hackathon 2026 ($2,000,000 in prizes)
π Opportunity Desk
π· GRANT
β³ Deadline: 2026-08-17
π Global, Worldwide
Deadline: August 17, 2026 Applications are open for the Build with Gemini XPRIZE Hackathon 2026. The $2 million global hackathon challenges developers worldwide to use artificial intelligence tools to build solutions across five categories: Education & Human Potential, Entrepreneurship & Job Creation, Small Business Services, Money & Financial Access, and Professional Services. The Build with [...
β Who can apply
πͺπΉ Open to Ethiopian applicants
#Grant #Ethiopia #Opportunities #Business #Education #OpportunityDesk
π Opportunity Desk
π· GRANT
β³ Deadline: 2026-08-17
π Global, Worldwide
Deadline: August 17, 2026 Applications are open for the Build with Gemini XPRIZE Hackathon 2026. The $2 million global hackathon challenges developers worldwide to use artificial intelligence tools to build solutions across five categories: Education & Human Potential, Entrepreneurship & Job Creation, Small Business Services, Money & Financial Access, and Professional Services. The Build with [...
β Who can apply
πͺπΉ Open to Ethiopian applicants
#Grant #Ethiopia #Opportunities #Business #Education #OpportunityDesk