PDF Experts in Germany

Part-time assignment since 01.09.2026.

• Operational support for fixed asset and investment accounting
• Support with the accounting treatment of assets under construction
• Processing of related accounting transactions in SAP R/3

System: SAP R/3

• Optimized the data architecture (Medallion) to better decouple processing steps and improve transparency and reproducibility

• Ensured the technical quality of data processing in Databricks by introducing schema enforcement, data quality checks, and a structured data architecture

• Pipeline orchestration with Azure Data Factory

• Improved and automated the development and deployment process by integrating Git and GitHub Actions

• Technical leadership of the Data Engineering team (3 employees)

• Led workshops to optimize and stabilize the data platform and development process

• Collected and prioritized new requests and maintained the product backlog

• Technologies: Microsoft Azure (Data Lake, Data Factory), Databricks, Apache Spark (PySpark), Python, SQL, Git, Confluence, Power BI, Power Apps, Dataverse, MS Sharepoint, Mural

• Functional requirements management for DIVID, the Umweltbundesamt platform for administering the single-use plastics fund. Focus: entitled parties, sub-entitled parties, and benefit reports.
• Consolidation and analysis of specifications, Jira, Confluence, functional concepts, and UX designs. Independent structuring and delimitation of requirements, as well as identification of contradictions, gaps, and dependencies.
• Creation and quality assurance of implementable user stories with clear, testable acceptance criteria. Preparation of decision alternatives and coordination with business, development, UX, and QA through team refinement.
• Conducting manual and technically assisted accessibility tests. Coordinating findings with QA and the Product Owner for further planning and implementation.
• Result: more consistent requirements, clearly defined processes and stories, and transparent open decisions. Separation of assumptions from confirmed requirements; better basis for implementation, refinements, and QA test cases.

Tools used: Jira, Confluence

Evidence-led health intelligence platform turning published research into personal health protocols. It scores 430 habits, foods, and supplements against the studies behind them, and moves the score when the evidence moves. Built solo.

• Built the daily ingestion pipeline across PubMed, bioRxiv, and medRxiv: 43,000+ papers from 3,400+ journals processed into 230,000+ typed evidence claims, each one traceable back to the study it came from.
• Designed the six-factor evidence scoring model and the public changelog behind it, so no recommendation ever appears without the papers underneath it. 23,000+ grade changes recorded and explained to date.
• Shipped an entity information model connecting every intervention to its mechanisms, biomarkers, and outcomes: 118 biomarkers with region-specific reference ranges, 77 mechanisms, 32 graded outcomes.
• Built the personalisation layer: blood panel ingestion that reads lab PDFs with a vision model and corrects results for draw time against the user's wake anchor, plus Oura, WHOOP, and Withings integration for daily readiness context.
• Operate eleven specialised review agents over the corpus and codebase, covering paper curation, retrieval quality, health-claim compliance across EU and US regimes, and security.
• Shipped the Evidence Assistant, a RAG assistant that answers from the claim database and cites the underlying papers, plus a B2B practitioner tier, an Expo React Native app, and localisation across 3 languages and 7 markets.

Stack: Next.js 16, TypeScript, Supabase, pgvector, Anthropic Claude, Vercel, DeepInfra.

AI-supported document processing and inventory management integration

Design and development of an AI-supported application for the automated processing of delivery and invoice documents, connected to SelectLine ERP. Documents are analyzed using AI, matched with orders and line items, and prepared for posting goods receipts.

IMPACT:

• Automated extraction of structured order, delivery and invoice data from PDF and image documents
• Automatic and manual mapping of documents to orders and order line items
• Integration with SelectLine ERP for order import, status synchronization and goods receipt postings
• Traceable processing through separate analysis, mapping and posting processes as well as technical logging
• Development of a containerized end-to-end architecture with AI analysis, workflow automation and relational data storage

KEYWORDS: AI, document analysis, OpenAI, n8n, SelectLine ERP, FastAPI, Python, JavaScript, MariaDB, Docker, REST API, PDF, OCR, mapping, inventory management, goods receipt, workflow automation

Multi-Agent Document-Generation Platform

Problem:

Turn a single prompt into a well-researched, structured and cited document, while keeping the user in control: every stage can be reviewed, corrected or revisited before the system moves on.

Architecture:

A multi-agent system orchestrated by a central controller. More than 20 specialised pydantic-ai subagents (understanding, structuring, planning, research, writing, assembly, cross-reference check, glossary, citation, reference table and more) pass results forward through a shared plan persisted in Postgres (asyncpg); a runner layer dispatches each task by name to its agent, and mechanical steps such as formatting, table of contents and list of figures are deterministic code instead of LLM calls. Provider (Anthropic or OpenAI) is a configuration setting, and responses stream over Server-Sent Events.

Hosting & Infrastructure:

Stateless container run as a Scaleway Serverless Container on a private network with managed PostgreSQL, object storage for finished documents and assets, and centralized log shipping (Loki / Scaleway Cockpit), provisioned with OpenTofu modules and deployed through GitLab CI; usage is metered per API key.

My Contribution & Key Decisions:

• Central decision controller: a satisfaction classifier turns the user's free-text feedback into a verdict (keep working, advance, cancel or revise an earlier task), and a middleware applies it to the plan. Plan progression supports jumping back to earlier tasks and re-running the stages after them.
• Research and writing fan-out: planning produces a plan from which a research-design agent generates individual research tasks and a writing-design agent generates individual writing tasks, each linked to the research it needs. Every task runs in its own dedicated agent with its own conversation history.
• Self-review and human-in-the-loop: after each step a critical-thinking agent reviews the output; the user can accept, refine, or ask questions about a critique before the plan advances.
• MCP integrations: two remote MCP servers, web search and internal document search, connected through pydantic-ai with retries and long-running tool-call timeouts. Classifier agents decide per task whether web search is needed, and a small model condenses raw search results into structured findings before they reach the writing agents.

Usage Metering:

Clients authenticate with an API key validated against a key-management service; consumption is charged in credits calculated from token usage per model plus a markup, and exhausted quota is answered with HTTP 429.

• Administration and further development of the M365 environment, including Teams, Exchange Online, Intune, and SharePoint Online
• Development of dashboards for performance monitoring with Microsoft Power BI
• Implementation of user requirements and processes with low-code/no-code technologies (Power Apps, Power Automate)
• Use of Jira and Azure DevOps to plan and implement user stories in an agile project context
• Participation in daily stand-ups, sprint planning, sprint reviews, and retrospectives as the technical point of contact
• Planning and implementation of security policies using Azure Conditional Access and multi-factor authentication
• Configuration and deployment of clients and applications via MECM (SCCM) and Baramundi

• Modernization of a large-scale Java enterprise application*

The project is modernizing a complex enterprise application that has grown over many years. The existing Spring-based legacy system runs on Java 8, OSGi, and Eclipse RCP and is being gradually migrated to a modern, maintainable architecture with Java 25 and Quarkus.

Marcus works on analysis, architecture, refactoring, and implementation. One focus is on untangling historically grown structures and dependencies and on building a clean, sustainable Java and Quarkus technology stack.

Tools & technologies: Java 8, Java 25, Quarkus, Hibernate ORM with Panache, EclipseLink, OSGi, Eclipse RCP, Maven, JUnit, Mockito, REST, JSON, Git, Eclipse IDE, IntelliJ IDEA Ultimate, Jira, Confluence

• Client: Prime Force Group GmbH

Technologies used: Java 25, Spring Boot 4, MapStruct, JSpecify, PostgreSQL, Redis, Liquibase, REST/OpenAPI, Apache Kafka, Apache Solr, OpenID Connect via IronGate/Keycloak, SAP Customer Data Cloud, JUnit, Testcontainers, Karate, Playwright, GitLab monorepo with CI/CD, Jenkins, JFrog Artifactory, FluxCD, Docker, Kubernetes on Azure, OpenTelemetry, arc42, Jira, Confluence

The Team Management Center is the new central platform of the DFB for planning, managing, and carrying out team activities for the national teams - from squad selection and training camps to communication with players, clubs, and legal guardians. The platform is designed for multi-tenancy for the DFB and regional associations; player, club, and master data are intentionally not copied, but connected at runtime via the DFBnet APIs.

I have been involved in the project continuously since the architecture and concept phase (Sprint 0) and work in a distributed Scrum team in two-week sprints. In addition to implementation, my focus is on architecture alignment, connecting the DFBnet interfaces, as well as code reviews and test automation as quality assurance in the team.

Focus areas:

• Development and implementation of the multi-tenancy concept (tenant model for the DFB and regional associations), including data model, access layer, and Liquibase migrations.
• Design of the person service and the search concept based on Apache Solr.
• Integration of the DFBnet APIs (player, person, and club search, club data), including authentication and synchronous master data synchronization.
• Hardening the integration through resilience patterns: separate read timeouts for each search path, correction of circuit breaker counting, limiting parallel requests, and a club cache to reduce load on the external system.
• Development of self-service endpoints for players (own activities, activity details, games), including an access concept for participants, as well as person documents and file uploads.
• Standardization of API design: OpenAPI annotations, nullability model via a custom ModelConverter, JSpecify migration of the DTOs, and documented API guidelines.
• Build and maintenance of Karate-based API and integration tests, integration tests with Testcontainers, test guidelines, and bug triage from the integration and reference environments.
• Code reviews via merge requests, architecture documentation according to arc42, and architecture decisions (ADRs) in Confluence.
• Automated deployment to the integration and reference environments, analysis of login and OIDC issues in combination with IronGate.

Status: ongoing - as of 08/2026 in Sprint 17, around 940 person hours worked; testable delivery to the integration environment every two weeks.

Strategic Marketing / New Customer Acquisition (B2B):

• Market and competitor analysis: Recorded competitors, developed evaluation matrices, outlined the provider landscape with potential customers, competitors and partners
• Designed a cross-media, personalized multi-touch campaign
• Developed the multi-touch strategy in various multi-stage versions, determined touchpoints, target companies and target persons, developed ideal customer profile and contact ideas
• Designed all individual measures: Developed concepts for letters, postcards, landing pages, e-mails, phone calls and downloads
• Visualized the process logic of the contact sequences (flowchart)
• Described the distinction from “Account-Based Marketing”
• Implementation planning: Researched providers for packaging, shipping and landing page software, negotiated terms, recommended tools and partners

Social Media Marketing:

• Goal categorization, definition and recommendations for action
• Review and selection of channels
• Design of channel management
• Developed an 11-week editorial plan (Instagram & LinkedIn)
• Developed, implemented and published image and video posts
• Campaign launch, ongoing monitoring and optimization

Search Engine Optimization (SEO):

• Website audit with indexing check
• Ranking, backlink, traffic and competitor analyses
• Analyzed search and user behavior
• Consulting on SEO risks during relaunches with domain changes
• Keyword research and recommendations for action

• Designed and developed a multi-tenant SaaS platform enabling organizations to build their own knowledge bases and chat with brand-customized AI assistants (white-label approach with dynamic branding per organization).
• Implemented a scalable RAG architecture with a GPT-4o tool-use loop, hybrid semantic search, and strict tenant isolation at database and search index level.
• Built persistent, project-like chat sessions including a streaming API (SSE), multilingual support, and speech input/output (STT/TTS).
• Delivered the cloud infrastructure as Infrastructure-as-Code, fully automated per-customer CI/CD pipelines, and an onboarding process for new tenants.

Technologies used: Python, FastAPI, Pydantic (v2 noted), Next.js, React, TypeScript, Tailwind CSS, OpenAI / LLMs (GPT-4o), Azure AI Search, Cosmos DB, Azure Blob Storage, Azure Cognitive Services Speech, Azure App Service, Azure Container Registry, Retrieval-Augmented Generation (RAG), Server-Sent Events (SSE), Docker, Terraform, GitHub Actions, REST, OpenID Connect (OIDC), Multi-Tenancy

Designed and developed an Azure-based AI and data platform transforming distributed project documents into searchable, structured knowledge for an enterprise AI assistant. Combined incremental document processing, hybrid keyword/vector retrieval, knowledge graph extraction, and conversational AI for Microsoft Teams and Microsoft 365 Copilot. Implemented Python document-processing pipelines for PDF, Word, Outlook and text files; Azure Cosmos DB lifecycle management; Azure AI Search; Azure AI Foundry and Azure OpenAI; identity-aware retrieval with Microsoft Entra ID; project authorization; a TypeScript Microsoft Agents SDK agent; Pulumi infrastructure; resilient scheduling, retry and diagnostic tooling. Technology stack: Python, TypeScript, FastAPI, Pydantic, Azure AI Foundry, Azure OpenAI, Azure AI Search, Azure Cosmos DB, Blob/Table Storage, Container Apps/Jobs, Service Bus, Key Vault, Entra ID, Docker, Pulumi, pytest, Microsoft Teams and Microsoft 365 Copilot.

AI Platform Backend – Senior Developer

Hired to design and build an AI platform backend from the ground up, including multi-provider LLM orchestration and real-time infrastructure for AI influencer personas at scale.

Tasks and Responsibilities

• Architected and implemented a multi-LLM orchestration layer with Semantic Kernel to integrate GPT-4 and other providers for core platform logic and AI influencer personas, reducing model-switching overhead by abstracting provider APIs behind a unified interface.
• Designed and developed a backend from the ground up in C# / .NET 10, including domain modeling with DDD, a versioned RESTful API layer, and cloud infrastructure on Azure.
• Built real-time chat infrastructure with Server-Sent Events (SSE), message persistence, and delivery guarantees for running AI influencer personas live at scale.
• Developed a media management service with cloud object storage integration for uploading and retrieving influencer-generated content.
• Created an integration and unit test suite with data seeding for reliable regression testing across all core platform flows, significantly reducing the production error rate.

Tools and Technologies: C#, .NET, ASP.NET Core, Python, TypeScript, MySQL, Semantic Kernel, EF Core, Minimal APIs, LLM Orchestration, Prompt Engineering, Agentic AI, Generative AI, AI-Assisted Engineering, Claude Code, GitHub Copilot, Google Gemini, OpenAI API, Ollama, Redis, Azure, Azure Container Apps, Azure Database for MySQL, Docker, GitHub Actions, Clean Architecture, Vertical Slice Architecture, CQRS, Domain-Driven Design, REST API, xUnit, Integration Testing, Unit Testing, Jira, Confluence, Scrum

• Adding/creating test specifications
• Performing regression tests
• Performing release tests
• Analyzing Jira tickets
• Identifying software defects and fixing them with C#