Large Language Model Experts in Dusseldorf

Project: Concept and implementation of an AI product for four business use cases

Project management of an AI project at a leading provider of large-scale catering and food services, where a production-ready AI product for four use cases was implemented together with an external development partner: automated briefings from CRM and document data, voice-based capture and structuring of reports, detection and merging of duplicates in master data, and data-based market analysis. A central focus was a privacy-compliant architecture that passed the internal IT security review and enabled productive use.

• Translating business requirements into clearly defined AI use cases with a clear product scope and clear value proposition
• Selecting and evaluating models and architecture options for text extraction, speech-to-text and context enrichment from business systems, including LLM integration, function calling and retrieval
• Designing and enforcing an architecture with European hosting, data minimization and masking of personal data as a prerequisite for approval
• Managing the interfaces between business, IT, IT security and the external development partner under restrictive data access conditions
• Coordinating with CIO and executive management on data access, risk assessment and approval decisions
• Preparing the transition into productive use

• Acquires new B2B customers for an AI SaaS startup in the public tendering space and books product demos with IT decision-makers.
• Qualifies target customers based on a defined ideal customer profile, including discovery, needs analysis, and objection handling.
• Builds domain knowledge in public procurement (EVB-IT, German and EU tender portals) for technical discussions at eye level.

PLANT-MY-TREE®-per-order

The application enables Shopify merchants to automatically place tree-planting orders for every incoming order. By integrating ecological contributions directly into the purchase process, the manual effort for tracking and billing reforestation initiatives is eliminated. The system increases transparency for end customers through real-time visualizations of the ecological impact directly in the storefront. The architecture is based on a modular monolith with Spring Boot in the backend and an integrated React app inside the Shopify admin area. The solution uses webhooks to capture order data in an event-driven way and integrates the weclapp ERP system for automated monthly invoicing. An app proxy mechanism provides dynamic statistics such as CO2 compensation and planted trees without any performance loss for the merchant shop.

Tasks:

• Design of the modular software architecture based on Spring Modulith to ensure high maintainability
• Development of the event-driven business logic for evaluating Shopify orders via webhooks
• Implementation of automated invoicing by connecting the weclapp REST API
• Building the frontend using React Router and Shopify App Bridge for native integration
• Design of the database model and implementation of the persistence layer with JPA/Hibernate and Prisma
• Integration of internationalization processes for global use in the frontend and email communication
• Automation of deployment processes using Docker and GitLab CI/CD

Project skills: Java 25, Spring Boot, Spring Security, Spring Modulith, Hibernate, JPA, REST API, PostgreSQL, Maven, Liquibase, React, TypeScript, React Router, Vite, Node.js, Prisma, Zod, Docker, Docker Compose, GitLab CI/CD, Shopify CLI, Shopify App Bridge, Polaris, weclapp, i18next, Lombok, Vitest

• Patent application: European patent drafting and prosecution
• LLM practicing: Developed LLM-based tools for automated patent data retrieval, applying Python scripting to accelerate technical reviews.

• Process optimization, process design
• Business partnering, target vision
• Teambuilding
• Workflow design
• Budget, MEC
• SAP S4/Hana, Board

• Conduct data management and statistical analysis for clinical studies on behalf of CROs.
• Guest lecturer at Ivancity University, Paris, specializing in data anonymization techniques and statistical disclosure control.
• Provide scientific and medical writing services for pharmaceutical companies.
• Perform optical mapping data analysis and develop software tools with a focus on algorithm optimization and technical support.

• CRM process and feature consulting
• AI infusion workshops to introduce AI apps for critical business processes
• Prototyping – Vibe coding with Lovable
• Development of several apps for global bid management in the CPQ, Salesforce, and SAP S/4HANA environment
• Requirements and process management
• Introduction of a new data governance model
• AI infusion – supporting apps and business processes with AI applications
• Portfolio management of AI ideas: identifying and selecting AI projects in bid management
• Prototyping AI projects with Lovable.dev
• Building an AI data foundation in Snowflake
• Blueprint for different business units and global rollout
• An agile prototype-first approach with design thinking and vibe coding to prototype all applications
• Presentation at a global conference on using AI in business and validating ideas with design thinking & vibe coding
• AI infusion workshops at global conferences for prototyping ideas with business stakeholders

• Development of the SmartHub platform for RepairX.io (iOS app & web)
• Development of an AI-powered (clinical decision support) patient management platform for the Malteser Hospital to provide care for uninsured patients
• Development of a retrieval-augmented generation (RAG) system for the intelligent processing of medical data for ASCEND gGmbH
• Design of an AI-powered system for emotion analysis of guests and development of AI agents for automated accounting and compliance checks
• Planning of the RepairX.io platform (circular economy) and management of a DAO Hyperledger blockchain system for NGOs
• Development of internal audit systems for AI ethics violations in healthcare (according to the EU AI Act)

• Founding member of a partnership for project and interim management in Berlin

• Built and launched the AI-powered “E-Invoice Corrector,” an intelligent system for validating and correcting invoices, using Python, FastAPI, and machine learning. The system is now live at Fintom8.
• Converted the Corrector into a fully functional API, published with Swagger documentation for easy access and integration by internal and external consumers.
• Designed, experimented with, and optimized advanced LLM prompts and meta-prompting strategies to improve automated reasoning, error correction, and decision-making in agent workflows.
• Wrapped and integrated existing APIs within the Google Agent Development Kit (ADK) framework to enhance automation capabilities and conversational AI workflows.
• Implemented comprehensive unit testing using pytest and unittest, and employed breakpoint debugging (VS Code, pdb) to ensure code reliability, maintainability, and smooth runtime execution.
• Utilized Pydantic and Tabulate for structured data validation, API schema management, and clear tabular data representation in testing and debugging workflows.
• Pursuing the Google Cloud Professional Certificate.

• Creation of business and technical concepts as well as support for enterprise architecture management as part of the migration to the Adesso standard insurance solutions and the renewal of surrounding systems

• Interface definitions

• Business analysis and creation of the business concepts for input management and sales

• Creation of architecture guidelines and processes

• Processing technical and business decisions in the architecture board

• Definition of service processes and system management aspects

• Creation of the technical target operating model

• Creation of an integration architecture for AI/LLMs from the providers Commasoft and Insiders

• CommaSoft (Alan) webpages and web services

• Insiders for Smartfix based on web services

• Creation of a question catalog for a security/regulatory check.

• Analysis of threats using threat modeling / STRIDE incl. creation of a set of measures to secure the systems

System environment: Jira, Confluence, Java, Enterprise Architect, VAIT/DORA, KritisV, GDPR, IT ServiceMgmt, Adesso Ecosphere and Insure Health, DoPIX and successor product, SmartFix, SmartFlow, Zabas, Kolumbus, Clarc archive, BiPRO, AI / LLM, MS Azure, AWS, TOM/FMO

• Development of an interactive prototype for the concept of an AI-supported Enterprise Architecture Management tool. The goal was to present complex relationships between IT systems, business processes, and departments in a way that is easy to understand and to support decision-making in the context of IT transformations.

• The prototype combined data-driven analyses with guided questions and interactive visualizations. A central part was the integration of a RAG process to provide domain-specific EAM knowledge in context. The focus was on quick idea validation, user-centered interaction design, and the technical feasibility of a scalable overall concept.

• Design, implementation, and validation of a RAG process for domain-specific EAM knowledge with Python, LangChain, and graph/vector databases (Neo4J, Milvus)

• Business and technical requirements analysis as well as definition of an MVP

• Development of the architecture and technology concept using Angular, Spring Boot, GraphQL, and Kubernetes

• Design of an interaction concept and creation of a brand style guide with Figma

• Development of interactive prototypes with Angular, Konva.js, and TypeScript

• Integration of CI/CD processes with GitLab CI/CD