Kamil Grymuza

I joined Crowdspring as a team lead (later promoted to CTO) with the goal of rebuilding this creative services marketplace from the ground up. My team and I built the current version from scratch in about 18 months and launched it successfully to over 300,000 users.

I was responsible for selecting the right technologies (Python, React.JS, and AWS), writing the core parts of the back-end code (including complex data migration), mentoring developers, and making sure the project was progressing at the right pace by defining tasks based on business requirements and managing the work.

I built a complete life insurance underwriting, quoting, and policy issuance subsystem with TypeScript and Python.

After speaking with multiple third parties and internal stakeholders, I gathered business requirements and designed the solution, utilizing the serverless offering from AWS (API Gateway, AWS Lambda, DynamoDB, DynamoDB Streams, and AWS Step Functions) and implemented the code.

Due to the complexity of the insurance sales process, I implemented the system as a state machine whose transitions were triggered both by user actions but also external system events (such as SFTP file creation or responses received from third-party APIs).

Initially, I acted as an advisor and an architect who designed the back end, front end, and infrastructure (AWS).

Later, I helped as an engineer who wrote a significant portion of the back-end code (Python) and scaffolded the React Native project for the front end (using TypeScript).

Finally, I contributed as an interim CTO who organized the work (translated business requirements into work items and introduced an agile process), managed the development of the MVP (done by two engineers), and supported the engineers from the technical perspective as the most senior person on the project.

Thanks to my contributions, the MVP version of the Keenee app was successfully launched in Boston in December 2022.

As a Principal Engineer, I built an open banking integration service for 20,000 users, a key data source for the Track, and a few other features of the Wagestream mobile app.

I architected the data model to allow for efficient querying of data at scale (Wagestream had 700,000+ possible users of this feature on the platform at the time) and built a set of services to authenticate with banks, pull and keep the transaction data up to date, process it and finally expose to the front end via a RESTful API.

Because of the high concurrency inherent to this problem, I chose Node.js as the main language used in the implementation and deployed the services to AWS. PostgreSQL was used as a data store and was accessed directly with hand-optimized SQL queries.

As a crypto investor, I could not find a helpful yet reliable crypto portfolio tracker, so I decided to build my own.

I first evaluated the feasibility of building a more reliable competitor to Delta, Blockfolio, and Cointracker, using a couple of Python scripts and pulling the data from the Binance API, producing useful statistics, such as Profit & Loss, resulting from multiple transactions spread over time.

Encouraged by promising results, I built a working prototype mobile application in Vue (wrapped in a Cordova wrapper) and backed by a Python API, running in AWS over a few weekends. The back end pulled the data from the Binance REST API, which was also connected to a WebSocket to support live updates as soon as a new transaction occurred.

The project got to a point where I either had to get funding for it and dedicate all of my time to it or pause the development, and I chose the latter due to personal reasons.

I selected the technologies, architected, and built a working MVP mobile app over a few weekends.

The app was a fully functional prototype written in React (enclosed in a Cordova wrapper) and backed by a Python RESTful API deployed in the AWS cloud. It was available for iOS and Android.

Despite promising results from testing with early adopters, the project ran into legal restrictions concerning private parking space rental, and we decided to stop further work.