I'm a Law student who somehow has the ability to speak tech, and lots of it. Here's my CV.
I work on different things almost weekly. From class projects, machine learning advice for companies and trying to predict state examinations. Right now I'm looking into data science and data analysis of university student timetables to answer some cool questions about the student population. Find out about my highlights by clicking on the thumbnails below.
Kynect | kynect.ai
Kynect.ai is a scalable platform to provide students and universities with unique tools.
Examples include powering ConnectUL, through our product AKHILL + KNCT. We enable the service to read a student's raw timetable from organic sources like their institution's information portal, without any configuration, extra setup or software with their institution. ConnectUL has been able to follow the "zero stakeholders" model thus so far, ensuring freedom to grow while not relying on third party support.
Our newest product in development: Rekount, a dedicated receipt management system with Machine Learning OCR functionality. Enables a team, a university society, startup, or those delegated with debit/credit cards to instantly submit receipts to prove purchases with their devices camera or from their photo library. Those in charge of finances can easily search through uploaded receipts by: User, Merchant, Amount, ID, Date or if the uploader needs a reimbursement. All this while ensuring information submitted is accurate: Is the right name of the vendor attached to the receipt? Does the amount reported match the receipt? You'll be the first to know and take action. Supports both iOS and Android with automatic and no-input-required updates.
Language / Technology
JSX / Node.js, React.js, Express.js, Fastify.js, Redis, GCP
SparkED | sparked.ie
SparkEd is a youth leadership programme for transition year students.
Based within Enactus UL, there are three elements to the programme: volunteering, workshops and final group projects. Focus is on developing leadership qualities and decision making.
To create an experience that would help keep students engaged, we needed a website to compliment our workshops. To achieve, i'm developing facilities for students to log in, log hours, read news, learn new content and intereact with the workshops leaders.
PHP - Laravel, NGINX, Cloudflare, MailGun, Node.JS
Projected users will be 2000+, with on-going development of international franchising.
Guess the LC | guesstheleavingcert.ie
Guess the leaving cert is an on-going project into studying the nature of the predictability of the leaving cert. Using multivariate linear regression, and sampling we'll attempt to answer the question: can we guess the leaving cert?
The project comes from a place of questioning the legitimacy of the leaving cert in providing a fair realisation of "other" intelligences. With future aims to push pressure on reforms by nullification of randomness, let's see how far we can go for better change.
Project is currently on pause to work on ConnectUL.
R, R Studio, TensorFlow, Python, NGINX, Cloudflare
ConnectUL | RBBU 2019 Finalist
In technical development. Web-based UL timetable application to improve project group and social group timetable management and planning. Made primarily in response to the GDPR, which limited the ability of public searching of student timetables. Focus on user group flexiability and ease of timetable recommendations. Plans for "the not being told" information, like events that might clash with recommendations: by using intergration through public offerings of data sources of services like Snapchat, Facebook and others. We don't want the end user to have to argue with the timetable recommendations, we want to know where they might be going to go without even saying a word. Data collection is through clear and open terms of service. Planned data science aims for student population.
Top Ten Finalist of RBBU 2019, representing Ireland.
R, R Studio, PHP - Laravel, Node.js, TensorFlow, Python, NGINX, Cloudflare
RBBU Competition Entry - Redbull Basement University
Finalist | GECAS Aerovate Competition 2019
My paper and conceptual idea involved the real implication and focus areas within GECAS where machine learning could drive decision making within the organisation by transforming their historical data into data points to be used to train models in areas like customer bankruptcy prediction, deal permutation recommendations and tax strategies. While highlighting detailed practical issues and worker psychology in response to a potential implementation of such a business intelligence adaptation.
My project was accepted into the top five entries, and was invited to present to members of the senior leadership team in GECAS.