Who Am I ?
Certified Chartered AI Engineer Associate with 3+ years of experience executing data-driven solutions to increase utility of business use cases.
Passionate about building models that fix real world problems.
Seeking to deliver insights and implement action-oriented solutions to complex business problems at your organization.
What I Do ?
Perform thorough analysis of data with appropriate visualizations.
Perform modelling to deliver business insights.
Execute computer vision applications.
Develop Transformer based NLP applications.
What can I Offer ?
Defining the problem and understand the requirements to carve out a niche for your project. Clearly outline success KPIs.
Collect relevant data from various sources for business use-case. Transform variables and perform preprocessing to build database for further work.
EDA provides insights which can aid in identifying the optimal set of features, through deep analysis of variables.
Creation of new features that better represent the variables to the predictive models through feature importance, extraction, selection, and construction.
Training of a baseline model, and increasing complexity with algorithms. Perform hyperparameter tuning for optimal parameters alongside reduction of overfitting.
Productionize code through testing and cloud based deployment of entire pipeline to provide accurate results for end user.
Recent Posts?
The aim of this project is to forecast the price of gold as a time series.
The dataset provided contains the average price of gold on a monthly basis from 1950 till 2020. The column of gold price is measured as the price per ounce (Oz) in USD.
Prophet was used as it is a procedure for forecasting time series data based on an additive model where non-linear trends are fit with yearly, weekly, and daily seasonality, plus holiday effects. It works best with time series that have strong seasonal effects and several seasons of historical data. The forecasting was made for the year 2021 and an upward trend was predicted by golden-prophet.
Code can be found here.
Read more
This project aims to extract keywords from a set of LinkedIn job postings.
This is done via scraping from LinkedIn jobs search URL, https://www.linkedin.com/jobs. The user is encouraged to select relevant filters for a more precise search regarding specific jobs. Scraping is done using Selenium and the results are saved into a .xlsx file for further use. An automatic webpage should open and the program will automatically click, and copy the relevant information. This is done through going into ‘inspect’ and via finding elements by xpath, class_name, tag_name, css_selector from the webpage.
Regex is used on the job description column of the dataframe to clean it, a custom dataset of non essential words are removed from the text and words are then lemmatized using WordNetLemmatizer. A corpus is formed from the column and “english” stop words are removed and finally keyword extraction is done using TF-IDF or KeyBert. The user has the option to select which method to use from the command line.
Code can be found here.
Read more
This project describes the development and prediction model considerations of an object detection system for retail items.
Ideally, the goal is to automate parts of the retailing experience, namely the checkout procedure in a retail store. This is to be achieved by tracking and identifying objects that the customer puts in their physical shopping cart, and then using this information to build a “virtual shopping cart”, which would potentially remove the need for a cashier to scan the items and handle payments in a convenience store.
Code can be found here.
Read more
