Welcome to my portfolio!
Below you will find different examples of solutions to technical and business problems primarily in the blockchain DeFi area of business. There are both examples of python code, SQL code and Tableau work. Examples are kept short and simple, aiming to give you a broad overview of my experience and technical skills.
I prepared this project to show a data pipeline built with Airflow. It is a simple example of fetching some Uniswap swap data (WETH/USDC) of a single user from Dune using their API and comparing the prices with Crypto Compare’s API.
Here I have put some links to Dune queries I have done while working at Summer.fi. Please note that these might not be maintaned anymore and are owned by the oasis_app user on Dune.
This post will go through a couple of examples of the work I have done in Tableau. Charts are stripped of axis info to censor any sensitive information. Example are simple as it is difficult to show details of my work in a censored way.
Simple example of how to use the python snowflake-connector library to interact with a Snowflake database. Both reading and writing data.
In this example I will be extracting data from a GraphQL endpoint. Also, I will show the use of asyncronous execution, enabling the code to wait for the API server before continuing. Lastly, the code will show how I am only fetching only the newest data since the last run and appending this to the database.
Here I show how to extract a dataset created on Dune.com. I show both the python setup and the Dune SQL, which can be filtered with a paramenter, that can be controlled when querying the API with python.
Here I will show how to create a dataset from blockchain DeFi borrowing/lending transactions containing the state of all positions (or value locked) at any given point in time. We do this to be able to plot TVM for example (Total Value Managed), which is an important metric to keep track of.
This example shows how I am labeling users based on a number of criteria defined in a decision tree. This is to be able to determine whether the users are active or inactive week by week and makes it possible to chart the developement over time.
This post will have various small examples of the use of different python tools and frameworks, that I was not able to fit into the other examples. However, I find them worthy of showcasing.
This is an example of mapping different values in the results of an encoded questionnaire (4237 rows and 194 columns). In stead of the actualy brand name it only states brand_0014 for example. Also, the answers based on some scale only has numbers where we for example want “Adults over 18”, “Teenagers”, etc. in stead.
This project was about preparing weekly sales data from wholesalers selling various products. The main problem was that the data came as multiple datasets inside ONE csv file.