Sr. Machine Learning Engineer
Flowciety Gmbh
Keywords
Artificial Intelligence
Python (Programming Language)
Machine Learning
SQL Databases
Google Cloud
Scikit Learn
Java (Programming Language)
Application Programming Interfaces (APIs)
Amazon Web Services
Data Analysis
Skills
Vertex AI , LLMs, OpenAI, Google Cloud Platform, AWS, API, AI, data analytics, Data Strategy, Data Visualization, Django, Docker, Elastic Search, GPT3, ,Google Cloud Platform, Indexing, Information Extraction, Computer Science, Keras, legacy systems, Looker, Machine Learning, NumPy, python, SQL query, SQL, Scikit-Learn, scikit, Software Engineering, spaCy, Tensorflow, UML, web service. Python, Java, Web Engineering, Unit Testing, Blender, Logic X, Discord
Project history
06/2021
-
10/2021
Developed python API on top of Amazon Textract, DUS and Elastic Search for Key
Information Extraction on business PDF documents and contribute to larger product.
Contact : Tim Kiese - tim@flowciety.de
Information Extraction on business PDF documents and contribute to larger product.
Contact : Tim Kiese - tim@flowciety.de
09/2020
-
05/2021
Data Scientist
Saloodo; Remote
Conducting SQL query migrations by aggregating several versions of the platform
running in 50 countries, to 1 global platform.
Developed a Recommendation Engine using Python machine learning tools to assist
carrier managers to link shippers to carriers, optimizing price & suitability.
Assisted in Data Strategy for 2021 for Saloodo to manage 30000 Shippers and
connected them to 30000 Carriers.
Contact : Kathrin Koch - kathrin.koch@saloodo.com
running in 50 countries, to 1 global platform.
Developed a Recommendation Engine using Python machine learning tools to assist
carrier managers to link shippers to carriers, optimizing price & suitability.
Assisted in Data Strategy for 2021 for Saloodo to manage 30000 Shippers and
connected them to 30000 Carriers.
Contact : Kathrin Koch - kathrin.koch@saloodo.com
06/2016
-
04/2019
Machine Learning Engineer
Red Buffer
Developed satellite image segmentation & solar panel design projection web service,
formulated roof ridge prediction as an image-to-image translation problem and
developed a Generative Adversarial Network (pix2pix) achieving an IOU of 0.88 for
multi-class semantic segmentation of roofs, trees, chimneys, and shadows.
Formulated Roof ridge prediction & panel placement as image to image translation
problem, trained Conditional Generative Adversarial Networks (DCGAN) to
synthesize roof-ridges & panels form learned distributions.
Developed models on proprietary news and social media sentiment data for stock trading
system traded NY S&P500 with a Sharpe ratio of 3.32. (Time Series )
Developed hourly footfall prediction system using Poisson Regression for a theme
park located in 26 locations in the U.S with a MAPE of 8%. Modernising data from
legacy systems and recommended staffing for all locations based on the prediction,
optimised service cost vs delivery (Time Series Analytics, Digitalisation).
formulated roof ridge prediction as an image-to-image translation problem and
developed a Generative Adversarial Network (pix2pix) achieving an IOU of 0.88 for
multi-class semantic segmentation of roofs, trees, chimneys, and shadows.
Formulated Roof ridge prediction & panel placement as image to image translation
problem, trained Conditional Generative Adversarial Networks (DCGAN) to
synthesize roof-ridges & panels form learned distributions.
Developed models on proprietary news and social media sentiment data for stock trading
system traded NY S&P500 with a Sharpe ratio of 3.32. (Time Series )
Developed hourly footfall prediction system using Poisson Regression for a theme
park located in 26 locations in the U.S with a MAPE of 8%. Modernising data from
legacy systems and recommended staffing for all locations based on the prediction,
optimised service cost vs delivery (Time Series Analytics, Digitalisation).
Local Availability
Open to travel worldwide
Other
GitHub.com/irtza
