About

Hi, I am Sarem!

I am currently working on completing my PhD thesis where I am developing methods to inject expert knowledge into complex machine learning systems. My main tool at the moment are Gaussian Processes as they allow some pretty neat mathematical tricks.

If you are interested in this line of work, you can take a look at my Google Scholar profile.

Besides research, I like building actual solutions to real-world problems using the tools and techniques explained in this blog and my publications. I am also primarily working as a freelancer in the data sphere.

You can also find me on LinkedIn.


Freelancing References

Most of my past freelancing jobs have been related to finance or consulting in some way - both related to data science, machine learning and/or data engineering.

Besides freelancing, I also had some internships at well-known consulting firms. Despite the work being interesting and challenging, I am currently preferring the freedom that comes with freelancing.

If you are interested in working with me, feel free to reach out to me at sarem.seitz@gmail.com - below you can find some reference projects. Information is kept sparse as most jobs were done under strict NDAs.

Data engineering and machine learning for an investigative consulting firm

I was responsible for designing and developing most of the data platform on virtual machines. Most applications were implemented in containers - orchestration of pipelines via Apache Airflow. I also built some NLP pipelines.

Technologies (excerpt):

  • Apache Airflow
  • Apache Spark
  • Docker
  • Huggingface Transformers

Data engineering for a consulting firm

I was responsible for designing pipelines that connected an external tool with an AWS infrastructure which I was also developing. Additionally, I designed an Excel Web Add-In to provide frontend access to client-facing consultants.

Technologies (excerpt):

  • AWS
  • Flask
  • Docker

Algorithm design for an international publisher

Here, I was helping another contractor with customizing a boostin algorithm. I derived some key formulas and implemented the algorithm in Cython for performance reasons.

Technologies

  • Python
  • Cython
  • LightGBM

Development and research of portfolio optimization algorithms for a fintech firm

My main tasks were related to portfolio optimization algorithms. I was mostly developing in R where we needed to implement both the algorithms and the API for a microservice infrastructe. Also, I did some research on more advanced portfolio optimization algorithms

Technologies (excerpt):

  • R
  • R plumber
  • Docker

Machine learning development and data engineering for a hedge fund

The client was a well-known long-short hedge fund with >1 Bio. AUM. I was working on their data science team to provide relevant insights to the analysts.

Technologies (excerpt):

  • Python
  • Apache Spark
  • Databricks

Imprint/legal information

W├╝stenahorner Str. 9
96450 Coburg - Germany
Email: sarem.seitz@gmail.com
Phone: +49 176 30105864
VAT: DE310940213

Unless explicitly mentioned, all stock images are licensed under Adobe Stock Standard license