Software engineer with 4+ years of international experience across Deutsche Börse Group, Heraeus, and Tata Consultancy Services, specializing in bridging technical infrastructure and business needs within global enterprise environments.
- 🔭 Currently building production LLM applications — RAG pipelines, agentic AI systems, and REST APIs deployed to the cloud
- 🤖 Designed an autonomous AI agent integrating SharePoint and Jira via REST APIs, cutting manual documentation effort by 80%
- ⚙️ Engineered Python automation reducing manual data-processing effort by 70% in financial operations
- 🌐 Full-stack mindset: from data pipelines and cloud infrastructure to APIs and BI dashboards
- 📦 Product engineering approach — I gather requirements directly from stakeholders and ship end-to-end solutions
- 🎓 M.Eng. Information Technology — Frankfurt University of Applied Sciences
open_to:
roles: ["AI / LLM Application Engineer", "Automation Engineer", "Data Engineer", "Junior Software Developer"]
location: ["Frankfurt am Main", "Remote (Germany)"]
languages: ["English (C1)", "German (B1)"]
| Domain | Proficiency | Details |
|---|---|---|
| RAG Architecture | ▰▰▰▰▰ | FAISS vector search, LangChain chunking, OpenAI embeddings, production deployment |
| LLM Integration | ▰▰▰▰▰ | Groq, OpenAI, Ollama — REST APIs with temperature control & structured outputs |
| Agentic AI | ▰▰▰▰▱ | Autonomous agents integrating SharePoint & Jira for end-to-end workflow automation |
| API Engineering | ▰▰▰▰▰ | FastAPI, Pydantic validation, Swagger docs, CI/CD to Railway |
| Data Engineering | ▰▰▰▰▰ | ETL/ELT pipelines, event streaming (AWS SQS), data quality & monitoring |
| BI & Analytics | ▰▰▰▰▱ | Power BI (PL-300 certified), Microsoft Fabric, real-time KPI reporting |
📚 RAG Document Question-Answering System
Production-grade Retrieval-Augmented Generation system enabling intelligent semantic search and knowledge retrieval from complex document corpora.
| Stack | Python · FastAPI · LangChain · OpenAI Embeddings · FAISS · Groq LLM · Railway |
| Scale | Full document corpus ingestion — PDF loading, chunking, vector indexing |
| Performance | Low-latency inference via Groq LPU; FAISS similarity search over embedded chunks |
| Security | Environment-based secrets, secure API key management |
| Impact | End-to-end semantic Q&A over unstructured documents, live in production |
| Repository | Live API Docs ↗ |
Built the complete pipeline from raw PDF to answered question: LangChain document loading and chunking, OpenAI embedding generation, FAISS vector store retrieval, and Groq-powered answer synthesis — all exposed through a validated FastAPI REST interface with auto-generated Swagger documentation and GitHub CI/CD deployment.
💬 LLM Chat REST API
Production REST API wrapping a Groq-hosted LLM with configurable inference parameters and strict request validation.
| Stack | Python · FastAPI · Groq · Pydantic · Railway |
| Scale | Stateless API design, ready for horizontal scaling |
| Performance | Groq LPU inference with tunable temperature control |
| Security | Environment-based configuration, no hardcoded secrets |
| Impact | Reusable LLM backend powering downstream chat applications |
| Repository | Live API Docs ↗ |
Designed and deployed a clean, validated chat endpoint with Pydantic schemas, automated CI/CD from GitHub, and environment-driven configuration — a foundation service for LLM-powered products.
🤖 Autonomous Requirement-Intake AI Agent (Deutsche Börse Group)
Autonomous AI agent automating requirement intake and documentation workflows end-to-end in a financial services environment.
| Stack | Python · REST APIs · JSON · SharePoint · Jira |
| Scale | Enterprise-wide documentation workflows across financial operations teams |
| Performance | Reduced manual documentation effort by 80% |
| Security | Enterprise authentication within a regulated financial environment |
| Impact | End-to-end automation of requirement intake for capital-markets teams |
| Repository | Internal — Deutsche Börse Group |
Integrated SharePoint and Jira through REST APIs to autonomously capture, structure, and document requirements — eliminating repetitive manual work and accelerating delivery in a global financial services organization.
learning:
- Advanced Python & OOP design patterns
- LangChain agents & multi-step orchestration
- Production MLOps & LLM observability
building:
- AI-powered CV tailoring pipeline (Claude API + DOCX automation)
- Agentic workflow automation tools
exploring:
- Vector database optimization (FAISS, ChromaDB)
- Local LLM inference with Ollama
open_to:
- AI / LLM Application Engineering roles
- Data & Automation Engineering roles
- Collaboration on open-source LLM tooling
"Engineering is turning ambiguity into systems that ship."