1. Tell us your background and your role as a data engineer?
My name is Dennis Odoom, and I'm a data engineer at Fidelity Bank Ghana. I was part of the 2019 data engineering cohort for blossom Academy. After three months, I've learned new skills such as websites scraping, building pipelines, cloud warehouse, Google big query, Amazon technologies, and spark.
My main task as a data engineer is data integration, and in terms of data integration, one of the key aspects of data integration is what we call data warehousing, where you extract the data from a source and then stores it in a warehouse for analysis. After data Analysis comes Data visualization where you create dashboards in either power bi or tableau in Excel, to tell a story about your data.
2. The major difference between a Data Engineer and A Machine Learning engineer?
The main difference between a data engineer, and machine learning engineer comes to play with what they do with the data. Data engineers extract data, cleanse the data, and put it in the formats for the machine learning engineer to build models for predictions.
3. How is data science shaping the financial sector on the African continent?
Data is gold, so when you have data, you can make better decisions, and strategize. As long as you have data you would be able to do a lot of things like customer segmentation, fraud detection anomalies, you can strategize to make informed decisions.
4. Where does data assurance fit between a data engineer and the machine learning engineer?
The data Assurance sits more with the data engineer than the machine learning engineer.
5. Your advice for beginners starting in data engineering and Machine Learning?
Anybody starting needs to work more on projects because these projects would build you up. When I was at university, the math professor always used to tell me, you learn mathematics by doing mathematics, so you have to do it. The fortunate news is that Blossom Academy is a place where you work on projects every week.
6. How do you resolve Fraud Detention Anomalies?
In Fraud detection anomalies, you don't necessarily need to visualize the data at all times. Detection of any fraudulent activity is based on looking at the trend and knowing what differs from the trend.