making julia the standard

  • Julia IT-consultancy
  • Building full production-grade solutions, including data architecture, ml ops, dev ops, gdpr and user interface.
  • values the common good and supports respective academic or SDG projects.

Founded by Stephan Sahm

  • a full stack data science consultant
  • 10 years experience in data science
  • 5 years in consultancy

Our Network — Julia — VKB

Short term

  • PoC machine learning with Julia
  • PoC performance speed up with Julia

Mid term

  • forecasts of predictors
  • real-time analysis
  • simple performant alternative to Spark for Big-Data
  • individual ai dashboards

Long term

  • migrating from Python to Julia
  • migrating from Fortran to Julia
  • migrating from SAS to Julia
Which are the most beneficial use-cases to start with?
What are the pain-points with the current setup? — Julia — VKB

Scope short term

  • development by
  • data provided by VKB
  • infrastructure provided by VKB (if applicable)
  • 12 days
3 weeks (4 days a week)
1st week
2 days data setup
2 days first prototype
2nd week
2 days improving prototype
2 days dashboard
3rd week
2 days finalizing
documentation day
presentation & future

Scope mid/long term — support with all stages

  • requirements assessment
  • feasibility evaluation
  • dependency tracking
  • time & resource planning
  • training
  • development
  • deployment
  • operations
VKB expertise

  • neuroscience
  • automotive
  • telecommunication
  • media
  • energy
  • manufacturer & retailer
  • e-commerce
  • ...
  • probabilistic programming
  • mathematical optimization
  • differential equations
  • natural language processing
  • time series forecast
  • fraud/anomaly detection
  • recommender systems
  • error estimation
  • classification & regression
  • computer vision
  • big data & real time ETL pipelines
  • ...
  • DevOps & MLOps
  • AWS expertise
  • Azure & Google Cloud
  • infrastructure as code
  • building datalakes
  • setting up compute infrastructure
  • Kubernetes, Hadoop, HDFS, Slurm, ...
GDPR & compliance
  • traceable data flows
  • right to be forgotten
  • consent management
  • pseudonymization
  • anonymization
  • prevention of discrimination
  • ...

Julia benefits

Developed and incubated at MIT,
Julia was designed to solve the two-language-problem of applied mathematics

  • you need speed
  • you need an easy-to-use language
30x to 300x faster than Python
Python/R code can be easily rewritten in Julia and immediately runs massively faster.
Accelerations by a factor of 100 are typical.
More efficient development
Julia solves the two-language-problem in programming: Easier than Python and fast like C.
Your developers are able to quickly prototype and effortlessly transfer the results into production.
Better maintenance
Julia enables an unprecedented level of code sharing and reuse.
It will make your solution simpler and easier to maintain.
Built for data science
Applied mathematics is the main focus of Julia's ecosystem and community.
You want state-of-the-art machine learning, optimization, simulation or other advanced data processing? It is all there. offering

Building a Julia proof-of-concept
Not sure yet if you should use Julia?
The best way to find out is to start small, with a PoC of your choice.
Migrating parts to Julia helps you with all stages: requirements engineering, planning, Julia development, and training of your team.
Big data, real time & HPC
Is your data too large to be processed in time by a single machine? Or do you have streaming processes with realtime requirements? can provide you with solutions that scale in speed and size.
Individual training & consultancy offers trainings and assistance at all levels, whether at beginners stage, intermediate or advanced, for prototyping, analysis, reporting, or production use.
Every special request is welcome.

We also do high-performance-computing and general performance optimization.