Development of a platform for the use of microservice based machine learning models in operative IT environments and the business processes integrated in them.
• High-level design of the overall system
• Conceptual orchestration of the module classes
• Definition of internal and external interfaces
• Implementation of modules
• Design and implementation of a central repository
• Implementation of service encapsulation and orchestration
• Selection and preparation of test data sets
• Iteration management: definition, execution, evaluation
• Definition and evaluation of functional/non-functional requirements
• Development of technical conception derived from requirements
• Creation of data pipelines for structured and unstructured data (ETL)
• Data acquisition from web-based sources and third-party APIs
• Conception and creation of data models
• Data exploration and analysis
• Data preparation and feature engineering
• Creation, evaluation and integration of ML models
• Creation of frontends for different stakeholders
• Process automation and CI/CD
• Rapid prototyping
• Operation and maintenance, data quality assurance, code quality assurance
Techstack: SQL, python, pandas, sklearn, prophet, sktime, jupyter, atom, MS Visual Studio, beautifulsoup, selenium, bert, spacy, folium, rasterio, gdal, matplotlib, dash, pillow, geopandas, Azure Databricks, Azure HDInsight, Azure Blob Storage, Azure EH4Kafka, Azure Virtual Machines, Azure API Apps, Azure Functions, Azure Database for PostgreSQL, Azure Machine Learning, Azure logic Apps, Power BI, Amazon Redshift, Amazon EMR, Amazon S3, Amazon EC2, Amazon API Gateway, Amazon Lambda, Amazon Sagemaker