profile site
Michael Darko Ahwireng
Tamale, Ghana
Data Engineering
Python, Airflow, ScikitLearn, DVC, Kafka
SQL/NoSQL
Mysql, Postgresql, Mongodb
Web Development
Bootstrap, Django, Nodejs, JavaScript
GIS & Remote Sensing
ArcGIS, QGIS, Mapinfo, ENVI, PDAL, GDAL
About me
I am a Data Engineer with GIS specialization and experience in vegetation health and environmental monitoring using satellite images. I am experienced in Python, SQL, NoSQL, JavaScript/HTML/CSS, NodeJS, ArcGIS and QGIS. I have completed Big-Data projects with Kafka, and Airflow.
Looking forward to put my GIS skills, Data Engineering skills and Machine Learning knowledge to use by building solutions to problems.
Education
- University of Ghana ( 2019-2020 )
MSc.(GeoInformation Science)
- University of Ghana ( 2010-2014)
BSc.(Earth Science)
Built custom LinkedIn company scraper using Selenium and BeautifulSoup to save scraped data as a json
Used Python OOP to write a class to load saved json into pandas DataFrame
Used Python OOP to write a class for cleaning and transformation of loaded data
Built a pipeline to handle processes from extraction to loading
Used Airflow to orchestrate and save data into AWS S3 bucket
Worked on the display and loading of data into Streamlit dashboard
Worked on a team to design and build a database for the project.
Led in the acquisition and analysis of satellite images for the selection of best point for drilling boreholes .
Created maps using ArcGIS and QGIS for reconnaissance surveys and built projects
Planning and Leading geophysical surveys.
Supervised construction of 2 water systems and more than 100 boreholes
Projects
Flower image classifier
Used VGG as a pretrained network to build a new feed-forward network as a classifier using ReLU activations and dropout. The built model can be trained on any set of labeled images. However the classifier was trained to identify flower images
A blog web app
Used JavaScript (Express and Node.js) for building the server. The web app is hooked on MongoDB which is used for the backed and also an API for authenticating users. The web app is deployed on Heroku. link to app
Water flow Modeling
Used pdal and geopandas to build a custom module to interact with ept and lidar files to extract and transform data from cloud to build Digital Elevation Model and create Topographic Wetness Index from the DEM.
Corona virus tweet analysis
Used gensim as a base for natural language processing and streamlit as a means of displaying findings. Scikit-learn was used to build a tweet classifier as well
Causal inference on breast cancer data
Performed causal inference on the variables recorded to influence the diagnosis of tumor to be malignant or benign. Machine Learning was merged with causal inference by using XGBoost to extract the important feature on which the causal inference was made
User analytics in tel-co industry
Carried out EDA on telecom data usage by subscribers and classified users into groups using k-means algorithm using the elbow method to choose the optimal k-means.