Menu
SAMPLE PROJECTS
IBM PERSONALITY INSIGHTS TEXT SUMMARY IN PYTHON
In playing around with IBM Watson's Bluemix APIs, I noticed that they had their personality text summary written in Javascript, but not in Python. They had many other parts of their Personality Insights service written in Python.
So, I ported over the Javascript library to Python, though at the moment it still heavily relies on the Javascript library. I am hoping to do a much fuller port in the future.
The code for the wrapper, and documentation for it, is available at this repository: https://github.com/davidarida7/personality-text-summary-python.
So, I ported over the Javascript library to Python, though at the moment it still heavily relies on the Javascript library. I am hoping to do a much fuller port in the future.
The code for the wrapper, and documentation for it, is available at this repository: https://github.com/davidarida7/personality-text-summary-python.
SOCIAL MEDIA POSTING
Recently, my interest has sparked in playing around with social media APIs, namely those of Facebook/Instagram. So, I created two simple scripts using them.
The first, written in PHP, allows you to post to multiple social media platforms at the same time. Until APIs for other social media outlets are updated and not blocked out, the code allows you to post to Instagram and an Instagram story at the same time.
That code, and documentation for it, is available at this repository: https://github.com/davidarida7/social-media-post.
The second, written in Python, is a small application that allows you to send multiple users the same message individually on Facebook Messenger, because surprisingly that is not a feature.
That code, and documentation for it, is available at this repository: https://github.com/davidarida7/multipost.
The first, written in PHP, allows you to post to multiple social media platforms at the same time. Until APIs for other social media outlets are updated and not blocked out, the code allows you to post to Instagram and an Instagram story at the same time.
That code, and documentation for it, is available at this repository: https://github.com/davidarida7/social-media-post.
The second, written in Python, is a small application that allows you to send multiple users the same message individually on Facebook Messenger, because surprisingly that is not a feature.
That code, and documentation for it, is available at this repository: https://github.com/davidarida7/multipost.
MYDATA AND CLIENT PERSONAS
During the period of my internship with ADP (Automatic Data Processing, Inc.), I was tasked with working on a two-phase project.
The first phase was writing a Python application that would pull in existing client data and use external APIs to look for external client data available publicly as well as gain additional insights on said clients. Such external data and insights included the client's Twitter handle, company website, Google standard address, LinkedIn profile, Wikipedia page and summary, personality summary, keywords, logo link, and financial information. All this data is outputted to a csv file for use in the second phase.
The second phase was to build a web application that displayed this client data as well as other associate scorecards and dashboards. However, this had to be done in a clean and user-friendly manner, with a design that is fun to use to the end user, not the standard and complicated business design.
Since the code was written for ADP, it is against their policy to share it externally. However, I have included screenshots so that you can gain insights about what my project entailed.
The first set of screenshots are the command line interface that you view when running the Python program that prompts you to specify your exact specifications for the client data you want.
The second set of screenshots are screenshots of the web application. Notice how the web application grabs your logged in Windows user credentials and displays your name and rights on the top right corner and displays only the pages applicable to your persona.
The first phase was writing a Python application that would pull in existing client data and use external APIs to look for external client data available publicly as well as gain additional insights on said clients. Such external data and insights included the client's Twitter handle, company website, Google standard address, LinkedIn profile, Wikipedia page and summary, personality summary, keywords, logo link, and financial information. All this data is outputted to a csv file for use in the second phase.
The second phase was to build a web application that displayed this client data as well as other associate scorecards and dashboards. However, this had to be done in a clean and user-friendly manner, with a design that is fun to use to the end user, not the standard and complicated business design.
Since the code was written for ADP, it is against their policy to share it externally. However, I have included screenshots so that you can gain insights about what my project entailed.
The first set of screenshots are the command line interface that you view when running the Python program that prompts you to specify your exact specifications for the client data you want.
The second set of screenshots are screenshots of the web application. Notice how the web application grabs your logged in Windows user credentials and displays your name and rights on the top right corner and displays only the pages applicable to your persona.
ECROP APPLICATION
Our team of five Georgia Tech students has been given the opportunity to work on a project that will allow ophthalmologists from around the world to determine the Economic Model of Retinopathy of Prematurity (EcROP) for their specific country within a single application.
As it stands currently, many countries in the developing world simply do not screen for a preventable and treatable disorder that causes visual impairment and blindness in prematurely born children. Dr. Timothy Olsen, our client, and others conducted a study that found that screening these children would result in increased quality of life for those affected in an extremely cost effective manner.
The paper that was published on this topic focused on the United States and Mexico, and our client believes that if the model can be applied to other countries by native ophthalmologists then it would be easier to convince those in positions of authority (e.g. Health Ministers) to implement and fund the screenings. However, currently there is no easy way for these ophthalmologists to enter their own data and create a model for their country and our application would be that solution.
Our client and his team are currently using very complex Excel spreadsheets and other proprietary software to do their calculations, meaning that it is very prohibitive for others to do their own calculations with their own data. We are going to wrap our clients’ formulas in an easy to use application so that the barrier to entry would be very low. Ophthalmologists would then be able to use our application without having to do any calculations of their own.
Our team is very motivated to work on this solution because the problem itself seems both solvable and meaningful. A successful application could have profound effects on thousands of lives.
Our users will be ophthalmologists in the developing world who wish to create an EcROP model for their specific country. They will be able to use our application to enter data that they have collected and have the application do the calculations for them. They will then be able to export the calculations. They will need only minimal experience using a computer/mobile device to be able to enter their data and interact with the application. This should be the only category of users.
The users can expect to see a step by step process that asks for specific data that the user has collected and a description or notes about the data if applicable. There might be optional data that will be clearly marked. They can also expect that once the application is finished calculating their model, they will be able to export it to different potential formats.
The link to the web application is: http://ecrop-env.4zgda28xd7.us-east-1.elasticbeanstalk.com.
The source code is available here: https://github.com/dchudy3/Emory.
Below are screenshots of the working application.
As it stands currently, many countries in the developing world simply do not screen for a preventable and treatable disorder that causes visual impairment and blindness in prematurely born children. Dr. Timothy Olsen, our client, and others conducted a study that found that screening these children would result in increased quality of life for those affected in an extremely cost effective manner.
The paper that was published on this topic focused on the United States and Mexico, and our client believes that if the model can be applied to other countries by native ophthalmologists then it would be easier to convince those in positions of authority (e.g. Health Ministers) to implement and fund the screenings. However, currently there is no easy way for these ophthalmologists to enter their own data and create a model for their country and our application would be that solution.
Our client and his team are currently using very complex Excel spreadsheets and other proprietary software to do their calculations, meaning that it is very prohibitive for others to do their own calculations with their own data. We are going to wrap our clients’ formulas in an easy to use application so that the barrier to entry would be very low. Ophthalmologists would then be able to use our application without having to do any calculations of their own.
Our team is very motivated to work on this solution because the problem itself seems both solvable and meaningful. A successful application could have profound effects on thousands of lives.
Our users will be ophthalmologists in the developing world who wish to create an EcROP model for their specific country. They will be able to use our application to enter data that they have collected and have the application do the calculations for them. They will then be able to export the calculations. They will need only minimal experience using a computer/mobile device to be able to enter their data and interact with the application. This should be the only category of users.
The users can expect to see a step by step process that asks for specific data that the user has collected and a description or notes about the data if applicable. There might be optional data that will be clearly marked. They can also expect that once the application is finished calculating their model, they will be able to export it to different potential formats.
The link to the web application is: http://ecrop-env.4zgda28xd7.us-east-1.elasticbeanstalk.com.
The source code is available here: https://github.com/dchudy3/Emory.
Below are screenshots of the working application.
TAP GAME
Recently, I developed in iOS application development with Swift. I joined the Georgia Tech iOS club to further develop this interest. During the first meeting, we were given the basic skills to develop a single view game application. I further expanded upon it to make it a tapping competition game, in which two players, represented by "Hillary Clinton" and "Donald Trump" tap their respective parts of the screen for 10 seconds. Their scores are incremented and shown on the screen, and whoever taps the most wins.
This small application, which demonstrates some of my skills in Swift iOS application development, makes use of stack views, labels, working buttons, timers, alerts, background music, and an application icon.
My code is located here at this GitHub repository: https://github.com/davidarida7/TapGame.
Below are screenshots of the working application.
This small application, which demonstrates some of my skills in Swift iOS application development, makes use of stack views, labels, working buttons, timers, alerts, background music, and an application icon.
My code is located here at this GitHub repository: https://github.com/davidarida7/TapGame.
Below are screenshots of the working application.
CLEAN WATER CROWDSOURCING
During my Fall 2016 semester at Georgia Tech, I decided to join a project that supports the Georgia Tech Serve/Learn/Sustain program. Specifically, I am working on a team to create an application that supports finding clean drinking water and reporting on testing results. In many parts of the world, access to clean water is a severe problem. Many times, water is available, but people do not know where to get it. Other times, water may appear clean, but has unacceptable levels of contaminants.
Even in areas where there is a lack of clean water, the population has access to devices on which applications can run. Most of these devices are smartphones. However, the team I joined is creating a JavaFX application to provide the functionality.
With the application, people can report on locations where water is available. Users of the application can find the nearest water source or report on a new water source. Workers with test kits will also be able to report on contaminant levels. The application will provide historical graphs to show variations in water quality over time for a specific location.
Our code is located here at the team's GitHub repository: https://github.com/victoryexe/Victoryexe.
Below are screenshots of the working application.
Even in areas where there is a lack of clean water, the population has access to devices on which applications can run. Most of these devices are smartphones. However, the team I joined is creating a JavaFX application to provide the functionality.
With the application, people can report on locations where water is available. Users of the application can find the nearest water source or report on a new water source. Workers with test kits will also be able to report on contaminant levels. The application will provide historical graphs to show variations in water quality over time for a specific location.
Our code is located here at the team's GitHub repository: https://github.com/victoryexe/Victoryexe.
Below are screenshots of the working application.