Hi, I’m Ajinkya and am working as System Administrator and also leading a team of 9 members who are responsible for managing and maintaining Private Cloud Infra. Nowadays, for taking any decision it has become important to look through data for getting better results of decisions taken.
For every quarter, I need to pull data with a data size of thousands of rows for data analysis for team KPIs. Creating a report with meaningful information on its hours of manual efforts using multiple excels sheets. And is a lot of pain to analyse the data from multiple files and then combine all of those multiple data into a single file for the final result. Being a huge dataset, it takes more than 5-7 business days to clean and prepare data for data analysis. After analysing the data and finding the KPIs, most of the time it used to require revalidating the results for final analysis which was much tedious work and had become a mundane task.
I have been heavily using Pandas for data cleaning and analysis which has helped me to do an analysis of multiple files in a single window of Jupyter Notebook. With this, it has become easy to apply various techniques such as Inferential and Descriptive Statistics, Normalization, Probability and Time series analysis for pure analysis and combine data into a single dataset without being worried about changing the original files/ datasets. Matplotlib and Seborn for Data Visualization were used along with MS Powerpoint Slides for the final presentation to my stakeholders.
We worked with IT Governance Team to get KPIs matrices. While cleaning the data, we took assistance from stakeholders to validate the data. We also worked closely with Team Manager to find if any additional or in-depth analysis is required to find more information other than KPIs. The final draft of the reports was shown to Team Manager to help him to understand the various conclusion and way forward on the actions of the operation that needs to be taken. Earlier, data cleaning used to take 5-7 days now it can be achieved in 2-3 days. This has helped me to focus more on analysis rather than making sure if data is clean properly or now. Also, I’m able to provide more insights on operations and areas where the team performed or underperformed other than KPIs. We identified the areas to improve the different processes to improve the end user’s experience. We also found a way to Semi-Auto a few of the things which can speed up the overall process.
As this is done for my team currently, we are still in the process of showcasing what kind of wonders can be achieved using data science. This has helped me to increase efficiency in terms of time and effort. Which can be utilized for more time for operations and management operations.
With this, I am able to automate a few of the tasks for data analysis.
