Christian F.

Rapid prototyping of an agent-based version of an existing SaaS product for service planning in the gastronomy sector. The goal was to validate whether the core product could be transformed into an AI-driven, agent-based workflow, where shift plans are automatically generated based on historical planning data, weather forecasts, and employee availability – without manual planning in the web application. The prototype was piloted with key accounts of the SaaS company to gather qualitative validation data for a product roadmap decision.

Technologies:

Vercel AI SDK, TypeScript, React, Node.js, Custom LLM Tool Generation, REST API Integration, Agent Evals, LLM Orchestration

Migration of an unstable legacy system landscape with performance issues, system crashes, and high support effort to Microsoft Dynamics 365 on Azure without operational interruption. Phased transfer of prioritized system components with bidirectional parallel operation, while downstream functionalities temporarily remained in the legacy system. Responsibility for architecture, implementation, and technical leadership of a migration and integration team, including task planning, capacity planning, and technical guidance. Additionally, conception and implementation of cloud-based interfaces for connecting external input channels to Dynamics 365, as well as automated extraction and structured processing of document-based information. Focus on improved stability, secure and scalable cloud architecture, and robust integration and data processing processes.

Technologies:

Azure, Dynamics 365, C#, .NET, .NET Background Service, REST APIs, Azure App Services, Azure Functions, Azure ServiceBus, Azure SQL Server, Azure Storage, Azure Key Vault, Azure AI Document Intelligence, Managed Identities, Azure VNet, Subnet Design, Private Endpoints, Private DNS, Azure Monitor, Grafana, Terraform, DevOps CI/CD Pipelines, Self Hosted Agents, Windows Service, On-Premises Integration, Hybrid Architecture, Strangler Fig Pattern, Zero-Downtime Migration

Building a central cloud data platform on Microsoft Azure for daily collection, transformation, and provision of data from heterogeneous business and IT systems for onshore and offshore wind turbines. Extension and implementation of central ETL processes with Azure Functions for consolidation, validation, and transformation of source data according to the Medallion approach with Bronze, Silver, and Gold layers. Provision of curated data via an ASP.NET Core Web API with OData functionality for flexible querying and integration into business applications. Implementation of digitalization use cases along the value chain of renewable energies in collaboration with business units. Additionally, implementation of DevSecOps measures with RBAC, IP restrictions, Azure Security Center, and file-based security checks. The platform served as a technical enabler for digital twin scenarios, process optimization, and data-based decision support.

Technologies:

Azure, C#, .NET, ASP.NET Core Web API, OData, React, TypeScript, Azure API Management, Azure Functions, Azure Event Grid, Azure Data Factory, Azure Data Lake, Azure SQL, Cosmos DB, Azure Storage, Azure Key Vault, Managed Identities, Azure RBAC, Azure Security Center, Azure Container Registry, Azure Container Instances, ClamAV, Azure Monitor, Application Insights, ARM Templates, Bicep, Infrastructure as Code, DevOps CI/CD Pipelines, ETL, Medallion Architecture, Digital Twin, Data Platform, Data Pipelines, DevSecOps