Welcome to my humble Github account! You'll find in the public space:
A project about a system that automatically optimizes algorithmic financial strategies and test them using a paper trading API composed by:
- quotes-importer : Imports quotes and historical data from providers on which to backtest the strategies
- ta4j : Enriched open-source Java library for technical analysis (mainly used for its technical indicators)
- backtester : Backtests the algorithmic financial strategies on the quotes by basically bruteforcing different parameters (using dedicated Java annotations)
- tradingbot: A tradingbot using Akka that tests the resulting strategies in real conditions through a Paper trading api (Alpaca)
The components of the website letsfindamovie.com that uses a CQRS architecture and everything is reactive. Made of:
- movies-manager : Source of truth of the system that loads from csv files the movies, tags and ratings and serves them to all the caches
- converter-gateway: Converts rest calls to events propagated within the system to add tags and ratings to movies in all components, written in async Python with Aiohttp
- cache-tags-manager : reactive cache of tags using ElasticSearch as a datastore
- cache-movies-manager: reactive cache of movies using Redis as datastore and Micronaut as a Framework.
- cache-movies-manager-haskell: same cache but coded in Haskell, still needs to implement the update of movies in Redis
- cache-ratings-manager: reactive cache of ratings coded in Scala using Akka and Play as framework, with Eventstore as a datastore (event sourcing)
- movies-explorer-webapp: React/Redux/RxJs webapp to consult and edit movies/tags/ratings + to consult Grafana dashboards about the state of every components (technical + business) + a diagram of the architecture with the link to this repo
- k8: Scripts and configuration to start the system in Kubernete clusters locally and on Google Cloud
- docker: Scripts to load the components in Docker as a dev/test environment
images-classifier:
- A Flask service that downloads a pre-trained ML model using Pytorch and then use it to get the classification score of new images
- Some scripts using AWS SageMaker to train a ML model using Pytorch, save it on S3 and then deploy it on a SageMaker Endpoint
Redis-tester : Performance tests with Gatling of different Java frameworks and Redis clients for a project of GVentures LLC Service.
pypetter-docker: An example of how to make pyppeter run in docker
front-archetype: An archetype of a React/Redux webapp with Webpack and Babel to help kickstart frontend projects
I'm a freelancer always interested in new projects, my Upwork profile: https://www.upwork.com/freelancers/~01a7e4d1f5ebce66f4 and my linkedIn : https://www.linkedin.com/in/xavier-barrelet/