🔥 Product-focused Data Scientist 🔥
- Ad Tech (current)
- Gaming
- Non-Profit
I worked at small to large-scale companies, so I enjoy getting my hands dirty and solving complex data-driven problems. My passion lies in enabling product growth using data and statistical theories.
- End-to-End designing of company-wide product success measurement through constant hypothesis validation on user behaviors
- Production-level software & data tooling development. Projects involve data pipelining, object-oriented design/refactoring, integration-testing, and operational maintenance.
- Experience with petabyte-scale data handling techniques such as Spark
- Familiarity with Ad-tech domains; real-time bidding, incrementality testing, brand-lift, AB experimenting, etc
- Programming - Python, R, SQL (Snowflake, MySQL, Postgre), Shell, HTML, CSS
- Cloud Service - AWS, Docker, Kubernetes, Jenkins, Spark,
- Visualization - Tableau, Dash(python), Power BI, Excel
- Web - Heap, Google Analytics
- Version / Collaboration - Git, Wiki, JIRA, Confluence
| Repo | Description | Link |
|---|---|---|
| R Projects | R Portforlio in .R or .Rmd files by business topics (i.e. ML, DL, Text Mining, Time Series, etc) | Link |
| Python Projects | Python Portforlio mostly in Jupyter Notebooks by ML/DL framework (i.e. Pytorch, Tensorflow, Fast.ai, etc) | Link |
| Medium Blog | Tech medium blog that talks about data/ml | Link |
| Repo | Description | Link |
|---|---|---|
| AWS SageMaker in Production | End-to-End curated examples that show how to solve business problems using Amazon SageMaker and it's ML/DL algorithm. Mostly in Jupyter Notebook for easy accessibility | Link |
| PySpark | PySpark functions and utilities with Real-world Data examples. Can be used to build complete ETL process of data modeling | Link |
| Recommendation System | Production-level Implementations of Recommender System in Pytorch. Clone repo and start training by running 'main.py' | Link |
| Natural Language Processing (NLP) Examples | Full implementation examples of several Natural Language Processing methods in Python. Ordered in a personal level of complexity | Link |
| Repo | Description | Link |
|---|---|---|
| Bengali.AI Handwritten Grapheme Classification | Classify three constituent elements in the image, given the image of a handwritten Bengali grapheme | Link |
