Like computer programming and solving problems.
Want to be a tech entrepreneur.
Sept 2023 - present
Working as a Senior Data Engineer in the company’s data platform team.
a) “Standardize real-time subscriptions in Google Pubsub”. This work helps the BI team to increase the quality & speed of the business report generated for the bank management team. Estimated improvement expected is 20-30%.
b) “Migration of Large Database Tables from batched ingestion to real-time” as a technical lead guiding two technical members seamlessly without any down time for a 10 GB data.
Sep 2021-Sep 2023
a) Implemented ETL pipelines in java and python that ingest data from postgres and mysql databases to Google Bigquery OLAP data warehouse. Implemented real time data ingestion using debezium (https://debezium.io/) Reduced cost of resulting ingestion by 30%.
b) Carried out version upgrade of company’s in house data visualization tool Redash (https://redash.io/ ) developed in Python. Worked on system design of ETL pipelines for custom requirements of Business Intelligence team.
Project Intern
Mar 2021-Aug 2021
Executed simple user requests related to database on-boarding and table set-up from BI team.
Developed a machine learning model for predicting alarm status of a CNC machine manufacturing components based on the data provided by the company. Achieved an accuracy of 85% using Artificial Neural Network with 4 layers.
Worked on “Python based extraction of decision-making rules from excel based data”. Learnt the basics of statistics, data science and supervised machine learning techniques – linear regression, logistic regression and Linear SVM.
C,Java,Python
Competitive Programming
Microservice design, implementation and monitoring
Technical Documentation
Machine Learning
Docker and Kubernetes
https://www.udemy.com/certificate/UC-b944a644-1e83-4cab-84bd-5e697115544f/
Linux Administration
https://www.udemy.com/certificate/UC-d81c7992-277a-47ef-96a2-a2184d6d420b/
Use ChatGPT and Google Bard LLMs to accurately resolve Python bug tickets. The training data consists of a set of bug tickets and their resolutions by a human programmer. The test data contains a different set of bug tickets. Currently training the trainable layers of ChatGPT and Bard and tuning their parameters to improve accuracy on the test data.