Hi, my name is Florian Ebeling and I work as a freelance software developer. I was early developer and later Chief Software Architect for a successful startup until it was sold and have recently released a Mac application Keys. I also build great piñatas for my two sons.
Things I do
- Ruby on Rails development, Ruby C extensions
- Mac and iOS development: Objective-C (and C)
- Architectural evolution: make the system let you sleep well, or make it real fast, or both
Things I also do
- Machine Learning: classification
- Hadoop: make sense of data when it’s a lot
- Erlang: develop scalable and “forget about” systems
- Clojure development: not add even more Java
Some projects from last 4 years
Keys — Mac application for learning typing (Objective-C)
Keys is a Mac application with focus on high-quality UI, straight forward focus on the main purpose and without distraction, and with a data-driven learning review. Standalone application sold through the Mac App Store. (See the Website.)
Social Network for Fashion with Scalable Storage (Rails, Riak)
A social network focussed on fashion with scalability built in from the beginning. Deployment with Chef.
Stateless Advertising Targeting Server (Ruby and C)
The system was scaling quite smoothly to 14bn requests per month.
Event Logging Server written (Erlang)
The logging server uses simple connections to collect individual events which get stored as records. Syntactically they are free-form documents. Clients are decoupled from server failures and the system is very simple.
Server Notification System written (Erlang)
System for sending notifications to a cluster of servers, e.g. for configuration reload or log rotation.
Flexible Formatting Component for Profile Delivery (Ruby)
Ruby code is stored in a database and is used to create advertising profiles in any possible format for various customer needs. From an internal hash containing generic profile values any output can be created by combining and re-calculating values. Presentation using custom names or positional formats.
Automatic Classification (Ruby and C)
Take user profiles and classify them on a number of variable (age, gender, income, etc.) Online classification and offline training tools. Algorithm library integration by building Ruby C extension. Additional functionality: training and validation of statistical models.
rb-libsvm C extension (Ruby and C)
Open source C extension wrapping the most popular SVM implementation, LIBSVM, by Taiwan University researchers. (See on Github.)
Clojure bloom filter
Open source package implementing a Bloom filter, a statistical logical set implementation for very large data sets. (See on Github.)
Portmill CI server (Ruby and shell script)
A CI server for the MacPorts project. Written in shell script and Ruby with a Ruby on Rails web frontend. Open source prototype of an CI server for checking all changes to packages.
I have also lead development teams, done technical due diligence for an aquisition, presented at users groups, written product specifications (agile and otherwise), technical and user documentation, built up development infrastructure, automated deplloyment, among other things.