Tools: Pytorch, TensorFlow, Keras, NumPy, SciPy, Scikit-Learn, Matplotlib, OpenCV, Skimage, Pandas, Flask, Node.js, React, Next.js, Typescript
Artificial Intelligence Engineer, Swisscom, 2020 - present
- Developed a data-intensive AI application for a financial institution that utilized NLP and ML techniques to analyze customer feedback, resulting in a 30% reduction in fraud cases.
- Created an innovative data-driven solution for a retail company, analyzing customer sentiment from social media data to improve customer satisfaction and boost sales.
- Deployed a machine learning model for a marketing agency to predict customer churn, reducing churn rates by 25% and increasing customer lifetime value by 20%.
- Developed facial recognition model using TensorFlow's high-level API, Keras.
Fullstack Developer / Machine Learning Engineer, ABS Finance, 2016 - 2020
- Developed an automated scoring and recommendation system with large car sales data sets to provide optimal decision making to local used car dealers.
- Developed an image segmentation server with semantic segmentation with Deep learning techniques using Tensorflow, running at Amazon EC2.
- Developed an algorithm for generating a super-resolution image from a low-resolution image using a machine learning technique (Python and Tensorflow at Google Machine Learning Platform).
Fullstack Developer, AISA, 2014 - 2016
- Implemented a robust payment gateway integration for multiple merchants, utilizing XML and JSON APIs to securely process online transactions and provide real-time payment status updates.
- Created comprehensive API documentation and conducted training sessions for merchants, enabling them to seamlessly integrate their systems with our platform and leverage the full potential of our payment solutions.
- Collaborated with cross-functional teams to introduce an automated order fulfillment system, streamlining the order processing workflow and reducing manual errors by 50% .
Developer, Software AG, 2012 - 2014
- Bank Controlling and Forecasting/Budgeting Cube (Multidimensional Financial Controlling). System still in use now. Inspired by Quantum Mechanics concepts (linear operators in a Vector Space), and implemented in MDX.
- Risk analysis.
- Analytical CRM / behavioral Customer segmentation.