Freelancer
Keywords
Skills
Please refrain from cold calling, send me an email with project details first!
My expertise is in the design and development of data-intensive applications that use Machine Learning or NLP to solve problems in various domains.
- Fields: Machine Learning, Natural Language Processing, Information Retrieval, Recommender Systems
- ML/NLP: Keras, pyTorch, spaCy, Pandas, scikit-learn, gensim, numpy, Jupyter, matplotlib
- Backend: Java & Kotlin with Spring Boot, Python with Flask or FastAPI
- Databases: NoSQL (Elasticsearch, MongoDB, Neo4j), SQL (PostgreSQL)
- DevOps & Cloud: CI (Jenkins, Drone), Kubernetes, Docker, AWS (SageMaker, SQS, ECS, S3, Lambda, API Gateway, DynamoDB, RDS, Cloudwatch)
- Workflow Engines: Apache Airflow, Argo Workflows
- Frontend: JavaScript, Typescript, material-ui, React
- Others: Project Management, Technical Lead, Scrum, Kanban
My expertise is in the design and development of data-intensive applications that use Machine Learning or NLP to solve problems in various domains.
- Fields: Machine Learning, Natural Language Processing, Information Retrieval, Recommender Systems
- ML/NLP: Keras, pyTorch, spaCy, Pandas, scikit-learn, gensim, numpy, Jupyter, matplotlib
- Backend: Java & Kotlin with Spring Boot, Python with Flask or FastAPI
- Databases: NoSQL (Elasticsearch, MongoDB, Neo4j), SQL (PostgreSQL)
- DevOps & Cloud: CI (Jenkins, Drone), Kubernetes, Docker, AWS (SageMaker, SQS, ECS, S3, Lambda, API Gateway, DynamoDB, RDS, Cloudwatch)
- Workflow Engines: Apache Airflow, Argo Workflows
- Frontend: JavaScript, Typescript, material-ui, React
- Others: Project Management, Technical Lead, Scrum, Kanban
Project history
11/2019
-
Present
01/2022
-
06/2023
Machine Learning Engineer
Otto Group
(Consumer goods and retail, >10.000 employees)
Description
I build and evaluated machine learning models to predict product characteristics. One of my notable achievements includes developing a multi-output neural network that was successfully deployed in production, enriching millions of products. In addition, I extended and created several streaming services (stateful and stateless) and contributed to their functionality and scalability, allowing millions of products and predictions to be processed efficiently.
Summary
I build and evaluated machine learning models to predict product characteristics. One of my notable achievements includes developing a multi-output neural network that was successfully deployed in production, enriching millions of products. In addition, I extended and created several streaming services (stateful and stateless) and contributed to their functionality and scalability, allowing millions of products and predictions to be processed efficiently.
Summary
- Technology: pyTorch, Keras, Python, Java, Spring Boot, JPA, Kafka, AWS (SageMaker, PostgreSQL,
S3, Cloudformation, Cloudwatch) - Prototyping & evaluation of various ML models
- Development of a multi-output neural network that was successfully deployed in production, en- riching millions of products
- DB design and optimization for PostgreSQL databases
- Optimization of Kafka streaming infrastructure to keep up with growing amount of data to be processed
- Cost analysis & reduction of AWS services
- Improvements and extension of backend services
11/2020
-
08/2021
Senior Full-Stack Engineer
MetaCell
(Pharmaceuticals and medical technology, 10-50 employees)
Description
Collaboration on various projects in the neuroscience environment. Technologically, the applications in these projects often consisted of a Python backend and a React frontend deployed in a Kubernetes cluster.
Summary
Collaboration on various projects in the neuroscience environment. Technologically, the applications in these projects often consisted of a Python backend and a React frontend deployed in a Kubernetes cluster.
Summary
- Technology: Python, React, JavaScript, material-ui, Three.js, Kubernetes, Jupyter Hub, CodeFresh (CI)
- Besides product development also responsible for project management and sprint organization
- Building a deployment pipeline in Codefresh and configuration of the Helm Chart and Kubernetes Files
- Designed and implemented a feature to enable parallel simulations, consisting of UI views to con- figure simulations and backend logic to manage and execute simulations in multiple processes
- Developing new React Components and extending Redux Actions and Reducers to reflect a new UI flow
- Development of end-to-end tests using Jest and Puppeteer together with integration of the tests into the existing deployment pipeline
11/2019
-
04/2020
Machine Learning Engineer
SPRYLAB
(Media and Publishers, 10-50 employees)
Description
Continued at SPRYLAB to build a content intelligence platform that uses Machine Learning and NLP to automate processes in the publishing domain.
Summary
Continued at SPRYLAB to build a content intelligence platform that uses Machine Learning and NLP to automate processes in the publishing domain.
Summary
- Technology: Python 3, Flask, spaCy, Gensim, Pandas, sklearn, nltk, scikit-surprise, Jenkins Pipeline, Kubernetes, Docker, Elasticsearch, MongoDB, Apache Airflow, AWS S3, React, Type- script, Jupyter Notebook
- Implemented use cases include: topic discovery and seasonal topic detection, content recommendation, in-site link optimization
- ML aspects: unsupervised learning, text preprocessing, feature extraction, entity recognition, scoring weighting models, phrase detection, recommendation models (content-based, collaborative, hybrid)
- Requirements analysis via discussions with pilot customers and evaluation of surveys
- Conception of a service-based system architecture
- Building a multi-stage CI pipeline in Jenkins and defining the Kubernetes deployment files.
- Development of Python Flask services providing REST APIs used by the frontend or external services
- Feature engineering in Python for articles and phrases, involving libraries like nltk and spaCy to extend features (e.g. entities, POS)
- Development of various scoring models in Python and Elasticsearch that use weighted features to perform ranking by top-k retrieval
- Integration of topic modeling with scikit-learn as a basis for topic proposals. And integration of word2vec with gensim for query extension
- Extending the Elasticsearch based recommender system with collaborative filtering using scikit- surprise and hybridization of further recommendation systems via weighted scoring
- Management of sprint and roadmap
11/2014
-
10/2019
Software Developer
SPRYLAB
(Media and Publishers, 10-50 employees)
Worked on a diverse set of projects ranging from mobile apps to data-intensive backends such as a recommender system or a content-intelligence platform. In my last three projects I took over the role of the lead developer and collected experience as a project manager.
Please have a look at the Selected Projects section in my resume if you want to know more about the project details during this time.
Please have a look at the Selected Projects section in my resume if you want to know more about the project details during this time.
11/2013
-
10/2014
Quality Assurance
sprylab technologies GmbH
(Media and Publishers, 10-50 employees)
Testing of mobile applications and web frontends, creation of test catalogs, automated testing
Local Availability
Open to travel worldwide
