It looks like you are in United States. Would you like to go to https://onsiter.com/us/ instead?
Könnte verfügbar sein
(Aktualisiert vor 1+ Jahr)Microservice & Infrastructure Architect
Stockholms län, Sweden
Einheimische Svenska, English
- Managed Elasticsearch & Kafka at scale
- 2.5 Years experience of Kubernetes
- 5+ Years experience of Docker in production
Fähigkeiten (17)
Docker
Node.js
GCP
B2B SOFTWARE
Software Engineering
Python
JavaScript
TypeScript
FRONT-END
REDUX
Rails
Django
Ruby on Rails
RECRUITING
Erlang
Ruby
PHP
Berufserfahrung
Software Engineering Consultant
Github
2019-06 - 2019-08
Working with an undisclosed urban mobility client to improve the scalability of their order handling systems.
Technologies: Docker, TypeScript, Node.js, AWS, MySQL, Redis
Technologies: Docker, TypeScript, Node.js, AWS, MySQL, Redis
Software Engineering Consultant
Qualifier.AI
2019-04 - 2019-06
Qualifier.AI is an AI-based tool for automating the sourcing and qualification part of the sales process. At Qualifier I wor
ked on containerizing and making serverless important parts of the architecture.
Technologies: Ruby on Rails, Docker, Google Cloud Functions, Serverless Framework
ked on containerizing and making serverless important parts of the architecture.
Technologies: Ruby on Rails, Docker, Google Cloud Functions, Serverless Framework
CTO
Klira
2017-01 - 2019-04
Klira is Sweden's third largest broker for consumer loans. During my tenure we grew the team from zero to 18 tech team
members. My responsibilities included from architecture to architecture to compliance.
Our first challenge was in replacing the PHP legacy backend with a new micro-service architecture with Kafka as the mes
sage broker, all hosted on Kubernetes. A key aspect of this replacing the legacy system was in utilizing Kafka as an event
-log so that writes could be replicated and written to both systems.
We also decided to replace the front-end systems with a "micro-service"-based front-end architecture where each group of front-end use-cases, gets its own front-end code-base. This way, if a single front-end system turned out to be poorly
written it could easily be replaced.
Metrics and monitoring turned out to be important for us and the next step, was to implement a modern and scalable mo
nitoring solution that could be used by in a DevOps fashion by the developer's. I worked on setting up a modern metrics
stack based-on Prometheus and the accompanying JMX agent.
Technologies: Kotlin, Java, Spring Boot, Prometheus, Kafka, Grafana, MySQL/MariaDB, Docker, Kubernetes, Redux, Reac
t.JS, styled-components, Next.js, gatsby.js, Node.js Google Cloud Platform (GCP), AWS.
members. My responsibilities included from architecture to architecture to compliance.
Our first challenge was in replacing the PHP legacy backend with a new micro-service architecture with Kafka as the mes
sage broker, all hosted on Kubernetes. A key aspect of this replacing the legacy system was in utilizing Kafka as an event
-log so that writes could be replicated and written to both systems.
We also decided to replace the front-end systems with a "micro-service"-based front-end architecture where each group of front-end use-cases, gets its own front-end code-base. This way, if a single front-end system turned out to be poorly
written it could easily be replaced.
Metrics and monitoring turned out to be important for us and the next step, was to implement a modern and scalable mo
nitoring solution that could be used by in a DevOps fashion by the developer's. I worked on setting up a modern metrics
stack based-on Prometheus and the accompanying JMX agent.
Technologies: Kotlin, Java, Spring Boot, Prometheus, Kafka, Grafana, MySQL/MariaDB, Docker, Kubernetes, Redux, Reac
t.JS, styled-components, Next.js, gatsby.js, Node.js Google Cloud Platform (GCP), AWS.
acq. by Shopify
2018-01 - 2018-01
called themselves a "tumblr for e-commerce", the aim of the company was to make setting
up an online store easy and accessible to the masses. During my stay there I worked on the ARPU-team, focused on incr
easing the Average Revenue per User. As part of this I was involved in building key-parts of Tictail's app-store. A platfor
m that which 3rd parties to sell applications on the companies platform. We greatly valued quality and correctness both i n back-end and front-end parts of the application.
Technologies: AWS, MySQL, Python, flask-restful, JavaScript
up an online store easy and accessible to the masses. During my stay there I worked on the ARPU-team, focused on incr
easing the Average Revenue per User. As part of this I was involved in building key-parts of Tictail's app-store. A platfor
m that which 3rd parties to sell applications on the companies platform. We greatly valued quality and correctness both i n back-end and front-end parts of the application.
Technologies: AWS, MySQL, Python, flask-restful, JavaScript
VP Engineering
Aevy
2015-02 - 2017-01
Aevy built a search engine for finding candidates for recruiting. To do this Aevy aggregates data from various sources on the web such as Github, StackOverflow and Facebook.
My first task at Aevy was replacing the Ember-based front-end with a new front-end based on React.JS. As part of this w
e went from a semi-interactive app to one that is fully interactive and felt like a single page web app.
The next challenge was Dockerizing the Ruby on Rails application and the other services so that we could start using Doc
ker both in development and in production. After dockerizing the applications the natural next step was to introduce Ansi
ble so as to replace manual operations for the infrastructure.
The next project was replacing the basic queuing/worker system based on Redis/Sidekiq with RabbitMQ which offers imp
roved reliability, which allowed us to switch to a more event rather than task driven architecture.
As we grew we wanted to index more data and the next step scaling up our Elasticsearch cluster from 3 to 15 nodes (eac h with 64gb). The cluster was operated using Ansible and deployed to bare-metal servers.
After adding even more load our Postgres Cluster started performing poorly and we had to design a solution for this. To s
cale up beyond a single database node, allowed by postgres, we decided to switch to Cassandra for certain types of data, and I setup a cluster for this.
My final big project at Aevy was setting up a large scale email validation system that used SMTP to milions of email-addr
esses. To achieve the speed we wanted we needed to use a large number of connections and threads; so many in fact th at even Java could not handle that kind of load. For this reason I used Elixir, a variant of Erlang having better syntax.
Technologies: React.JS, Clojure, Ruby, Elixir, PostgreSQL, RabbitMQ, Ansible, Docker, Cassandra
My first task at Aevy was replacing the Ember-based front-end with a new front-end based on React.JS. As part of this w
e went from a semi-interactive app to one that is fully interactive and felt like a single page web app.
The next challenge was Dockerizing the Ruby on Rails application and the other services so that we could start using Doc
ker both in development and in production. After dockerizing the applications the natural next step was to introduce Ansi
ble so as to replace manual operations for the infrastructure.
The next project was replacing the basic queuing/worker system based on Redis/Sidekiq with RabbitMQ which offers imp
roved reliability, which allowed us to switch to a more event rather than task driven architecture.
As we grew we wanted to index more data and the next step scaling up our Elasticsearch cluster from 3 to 15 nodes (eac h with 64gb). The cluster was operated using Ansible and deployed to bare-metal servers.
After adding even more load our Postgres Cluster started performing poorly and we had to design a solution for this. To s
cale up beyond a single database node, allowed by postgres, we decided to switch to Cassandra for certain types of data, and I setup a cluster for this.
My final big project at Aevy was setting up a large scale email validation system that used SMTP to milions of email-addr
esses. To achieve the speed we wanted we needed to use a large number of connections and threads; so many in fact th at even Java could not handle that kind of load. For this reason I used Elixir, a variant of Erlang having better syntax.
Technologies: React.JS, Clojure, Ruby, Elixir, PostgreSQL, RabbitMQ, Ansible, Docker, Cassandra
Software Engineer
2014-09 - 2015-02
5Monkeys is web agency helping its customers build and implement their great ideas. I returned to 5Monkeys to continu
e working on various client projects. During this time I worked on an internal portal for a magazine publisher, and an e-co
mmerce storefront.
Technologies: Oracle, MySQL, Python
e working on various client projects. During this time I worked on an internal portal for a magazine publisher, and an e-co
mmerce storefront.
Technologies: Oracle, MySQL, Python
Software Engineering Contractor - Tictail
2014-05 - 2014-09
Software Engineer
2013-10 - 2014-05
5Monkeys is web agency helping its customers build and implement their great ideas. I worked with a variety of client pr
ojects all based on the Django + Python stack. One of the clients I worked with a was a peer to peer lending company, wh
ere I built the platform where lenders and borrowers could see each other's tenders.
Technologies: Python, Django, MySQL
ojects all based on the Django + Python stack. One of the clients I worked with a was a peer to peer lending company, wh
ere I built the platform where lenders and borrowers could see each other's tenders.
Technologies: Python, Django, MySQL
Software Engineering Contractor
AgFlow
2013-05 - 2013-10
At Agflow we were some of the earliest users of React outside of Facebook. As the only front-end developer I was respon
sible for building the front-end for our trading platform.
Technologies: React.JS, Go, Javascript
sible for building the front-end for our trading platform.
Technologies: React.JS, Go, Javascript
Akademischer Hintergrund
MSc.
2024-11 - 2024-11