• Tracking of users in the web and mobile application with Matomo Analytics
• Surface design and implementation in the web and mobile application
• Management and creation of content using the Contentful headless content management system
• Technologies: Headless CMS Contentful, JIRA, Confluence, AngularJS, NodeJS, Build-Pipeline Docker and Gitlab

In order to be able to process handwritten documents, the use of OCR tests (Optical Character Recognition) is necessary. Unstructured data is converted into structured data, for this purpose it must be recognized and checked with various regression test cases. For a clean reporting of the processed data in the OCR area, I was given the responsibility to set up a data analysis tool to clearly display the successfully processed documents. Skilja was used for this as a document understanding - deep learning platform.

• Module development for the creation of management reports using Microsoft Power BI
• Processing of unstructured data into structured data
• Testing of process flows for the classification of unprocessed documents
• Technologies: Skilja, OCR (Optical Character Recognition), Microsoft Power BI, Excel Pivot Tables, IBM Watson Visual
Recognition, Oracle DB, JIRA, Confluence

As a business analyst, I was also responsible for the implementation of the chatbot project for the automated processing of customer inquiries and for the further integration of the internal system used into the Facebook channel for the chatbot used

In order to be able to optimally measure marketing KPIs, a front-end reporting tool was required to visualize the marketing key figures

• Conception and implementation of a chatbot project for the automated processing of customer inquiries
• Content creation of the chatbot for different use cases in the workflow of the chatbot
• Creation of a reporting tool for the visualization of marketing metrics for the analytics area
• Technologies: IBM Watson Assistant, Genesys, IBM-Cloud, Jenkins, Cloudant, ElasticSearch, Kibana, VueJS, NodeJS, Java, JIRA, Confluence

As a developer for the Facebook channel, I was responsible for expanding the existing interface with additional functions from the Facebook Messenger API. For an excellent customer experience, you need an excellent customer workflow when handling customer concerns.

The following messaging elements played a particularly important role for an optimal user experience:

• Quick replies
• Templates
• Generic content
• Handover protocol
• Persistent menus

Technologies used: Java, Tomcat Service, jersey and jackson, IBM Cloud, Watson Natural Language Understanding / Natural Language Processing

BOHOFABLE was founded in 2018 with the intention of reviving a fine jewelry label to represent the long-forgotten Bohemian style. A variety of modern chic basics was offered, from necklaces to bracelets.

A large manufacturer of white labeling was selected for the products. Bohofable itself specialized in marketing on the Pinterest and Instagram channels to express the style of freedom, chic and bohemian.
During this time I was able to gain a lot of experience in the field of online marketing and e-commerce and managed and created all activities myself. From shop creation, marketing and branding to taking photos and videos on various social media channels.

After the initial development, the company was sold to a private Canadian investor who no longer continues the brand.

@bohofable

Chatbots and artificial intelligence are one of the hot topics in the tech scene. As a developer for a major customer, I helped digitize and automate the customer support process right from the start.

For this purpose, I helped to integrate such a project in the first step into the Facebook channel during the first analysis in order to also provide customers with an initial customer touchpoint for follow-up projects in omni channels.

The Watson framework from IBM, which was already awarded by "Jeopardy!" In 2011, was used as the NLP tool.

I was architecturally responsible for the integration for the Facebook Messenger service and designed and implemented the middleware for the Facebook channel between the Messenger service and the Watson API.