Stephan Fröde - NLP/LLM Chatbot

Q: Where is Stephan based?

Stephan is based in Wiesbaden, Germany .

Q: What languages does Stephan speak?

Stephan speaks the following languages: German (Advanced), English (Advanced) .

Q: How many years of experience does Stephan have?

Stephan has at least 30 years of experience. During this time, Stephan has worked in at least 24 different roles and for 16 different companies . The average length of individual experience is 1 year and 3 months . Note that Stephan may not have shared all experience and actually has more experience.

Q: What is Stephan's latest experience?

Stephan's most recent position is NLP/LLM Chatbot at Insurance .

Q: What companies has Stephan worked for in recent years?

In recent years, Stephan has worked for Insurance , Federal Criminal Police Office , Siemens , BMW Connected Vehicle , and Generative AI Blog & AWS Serverless Experiments .

Q: Which industries is Stephan most experienced in?

Stephan is most experienced in industries like Media, Entertainment and Publishing , Information Technology (IT) , and Telecommunication . Stephan also has some experience in Banking and Finance , Automotive , and Transportation and Logistics .

Q: Which business areas is Stephan most experienced in?

Stephan is most experienced in business areas like Product Development , Information Technology (IT) , and Business Intelligence . Stephan also has some experience in Research and Development (R&D) , Project Management , and Operations .

Q: Which industries has Stephan worked in recently?

Stephan has recently worked in industries like Information Technology (IT) , Media, Entertainment and Publishing , and Government and Public Administration .

Q: Which business areas has Stephan worked in recently?

Stephan has recently worked in business areas like Information Technology (IT) , Product Development , and Research and Development (R&D) .

Recommended expert

Wiesbaden, Germany

Experience

Oct 2024 - Apr 2025

7 months

NLP/LLM Chatbot

Insurance

Conceptualized and implemented an LLM-based case assistant (file assistant)
Selected and evaluated RAG methods; designed hybrid RAG information retrieval using Elasticsearch + embeddings
Built ingestion pipelines for multiple document formats; analyzed and aligned with source systems
Developed a Streamlit-based chatbot GUI and performed NLP-based causal chain analysis for regress cases
Evaluated analytical LLM methods; deployed via Jenkins to OpenStage

Mar 2024 - Present

2 years

AWS PoC / Content Analytics

Evaluated dashboard approaches and built a prototype dashboard in Streamlit using Google Search data and HTTP logs stored in AWS S3/Athena
Integrated web search and marketing/blog data; designed pipelines for image knowledge extraction and evaluation (LMM / Graph & Vector RAG)
Built an AWS Amplify-based service integrating AI-as-a-Service (Mistral, Replicate); data persisted in Neo4j with text-to-graph extraction
Implemented Graph RAG with Neo4j, and deployed with AWS CDK and GitLab CI/CD

Jun 2023 - Jun 2024

1 year 1 month

Knowledge Graph / NLP / AI

Federal Criminal Police Office

Worked in a security-cleared (SÜ2) and BSI-compliant environment
Explored and advised on Knowledge Graph use; designed & implemented pipelines with Neo4j
Raw data extraction from SQL dumps and mapping to entities; graph data model design
Advanced NLP: NER, topic & event extraction, language detection; extraction of complex relations using POLE (Persons, Objects, Locations, Events)
Applied graph algorithms (incoming degree, community detection – Louvain/triangle count) to identify relevant entities and relationships
Integrated GeoIP/domain extraction and URL evaluation; implemented streaming ingestion (Kafka)
Built an LLM/Chat Streamlit app deployed on Cloud Foundry on Kubernetes; integrated into BSI-compliant environment

May 2022 - Nov 2022

7 months

Corporate Web Portal

Siemens

Designed and implemented daily data pipelines using AWS Glue/PySpark and Athena; monitoring via CloudWatch
Built a GraphQL Apollo Server in AWS Lambda; integrated content items from OpenSearch and usage data from Adobe warehouse
Added an app for readability scoring; integrated an AWS SageMaker endpoint for text classification
Conducted feasibility studies, concept work, and stakeholder alignment; evaluated Athena Query Federation; managed costs and optimization

Mar 2021 - Mar 2022

1 year 1 month

Data Lake

BMW Connected Vehicle

Led development of large-scale AWS-based data pipelines for multiple datasets (Glue/PySpark, Athena, S3, Lambda, IAM, SQS/SNS, DynamoDB/DAX, Elasticsearch)
Processed up to 30+ billion rows; implemented automated restart & recovery procedures and unit tests
Architected daily/historic/housekeeping pipelines; performed production problem analysis and internal advisory
Performed cost analysis, derived measures, implemented cost savings, and monitored reductions
Refactored complex vehicle master data (7 datasets, ~35 tables; mid-level business logic; >1B rows/day)
Ran feasibility analyses, requirements refinement, and implementation concepts; authored documentation

Jan 2020 - Present

6 years 2 months

Project

Generative AI Blog & AWS Serverless Experiments

Built multiple serverless stacks in AWS using CDK and GitLab CI/CD (VPC/Serverless, Amplify, AppSync, Lambda, API Gateway, Cognito, Route 53, Glue, Neptune, SageMaker, Bedrock, S3)
Explored LLMs/NLP (OpenAI API, Hugging Face, Mistral, Replicate, Gemini, Claude); conducted experiments in automated image analysis & generation with VLM/LMM
Implemented AI data pipelines for knowledge extraction and enrichment (Image-to-Text-to-Knowledge); built chatbots combining image analysis with knowledge graph context
Developed iOS exploratory apps (Swift/SwiftUI) with API Gateway; results persisted in DynamoDB; Lambda/Python back ends

Jul 2017 - Aug 2020

3 years 2 months

Data Warehouse (Passenger Revenue)

Deutsche Bahn Vertrieb

Designed & implemented data model for passenger revenue (ticket sales and other services) considering AWS specifics (sort key, data distribution)
Built multi-stage Data Warehouse architecture: two-stage staging, Core (persistent business model), Analytics (optimized for SAP BO), and Output for file interfaces
Designed ETL/Data Pipelines: daily, historic reprocessing, housekeeping; Lambda/S3 events for job control; reload from raw data; processed ~1M rows/day
Implemented jobs with AWS Batch (Python, Docker) using SSM Parameter Store; integrated DynamoDB; limited Red Hat OpenShift on AWS
Managed build pipelines via GitLab CI/CD for >15 projects (dynamic parameterization, Docker images, build scripts, Python merge/deploy)
Conducted performance analysis and workshops with AWS Support; optimized Redshift WLM; ensured operations stability with weekly coordination & troubleshooting

Aug 2016 - Jun 2017

11 months

Inbound Logistics Costs

Opel/GM

Designed data model to compute inbound logistics costs based on freight rates, goods receipts, and master data
Calculated inbound costs per produced vehicle and per transport/shipment/material number; extensive analyses with SQL and R to derive business rules and error analysis
Acted across roles as Data Analyst, Business Analyst, and Data Engineer; coordinated interfaces and processes with logistics, network planning, and finance
Produced IT concept and handed over solution to the business unit for operational use

Apr 2016 - Jun 2016

3 months

Functional Data Analysis

BMW

Analyzed materials control and logistics planning; prioritized with stakeholders; investigated the SAP–DWH interface
Investigated Teradata SQL, ETL modules, views, and data models across source, inbound (SAP→DWH), and DWH layers; proposed solutions and documented results

Apr 2016 - Jun 2016

3 months

Production Data / Big Data

BMW

Authored interface specifications and conducted SQL-based data analysis; captured and analyzed requirements; documented & modeled the solution
Created the IT concept (architecture, interfaces, operations); supported tool selection for change-data-capture

Jan 2016 - Present

10 years 2 months

Side Project

Social VR Prototype

Prototyped collaborative VR space with messaging, user management, and real-time constraints
Explored scaling and availability of real-time services; created visual concept and 3D modeling

Sep 2014 - Feb 2015

6 months

Time Tracking

BMW

Prepared and completed go-live readiness and test management; authored operating manual
Analyzed deployment and operations issues

Apr 2014 - Mar 2016

2 years

Tableau / Teradata Introduction

BMW

Introduced Teradata/Tableau BI/DWH environment (~80 interfaces): architecture analysis, scaling/performance, and 7-year TCO across 6 alternatives
Supported high-level/functional/IT concepts, interface analysis, and coordination with external data suppliers
Advised on information security and personal data protection; initiated and automated operations process; managed change and incidents
Coordinated testing (test cases & monitoring); led code reviews, data quality analyses, and remediation proposals

Apr 2014 - Feb 2015

11 months

Strategic Reporting – Pre-Series

BMW

Coordinated changes and authored operating handbook for go-live readiness & operations enablement
Analyzed deployment/operations issues and documented operational measures
Worked with SAP BW (process chains, DSOs)

Mar 2013 - Jun 2013

4 months

Special Equipment Planning

BMW

Produced high-level and functional concepts: process analysis & description, requirements capture & analysis, process modeling
Translated requirements and process models into user guidance and data stewardship workflows

Feb 2013 - Mar 2014

1 year 2 months

Warranty Cost Controlling

BMW

Analyzed processes and IT architecture; ran performance tests; captured functional requirements (costs, quality, provisions)
Proposed BI/analytics solutions and a new architecture; delivered prototypes in QlikView, SPSS Modeler, and Java EJB

Jan 2013 - Feb 2013

2 months

BI in Quality Assurance

BMW

Analyzed data model and supply of a QlikView application in QA; built UI and analytical requirements in QlikView 11 with VBScript components
Implemented country-specific data views based on user (NTUser)

Jan 2012 - Present

14 years 2 months

Hybrid

Hybrid AI Storytelling Platform

Saganode

Designed and built a hybrid AI storytelling/world-building platform combining knowledge graphs, LLMs, and custom entropy-shaping heuristics
Implemented multi-layer graph spaces (temporal, causal, conceptual) and Graph RAG; integrated generative services for text, image, and music
Developed React-based graph explorer for universe/saga/story navigation
Managed security and deployment: AWS AppSync/Amplify back ends; REST/GraphQL endpoints; CI/CD; AWS CDK; Cognito/IAM

Jan 2012 - Dec 2020

9 years

Recommendation Engine

Obylo

Explored graph database approaches and semantic BI on top of Neo4j; built ETL for graph ingestion; researched scaling/latency/IO constraints
Developed early social VR experiments using Unity/Blender with Akka/C#; built search and recommendation components for complex data

Jan 2009 - Dec 2012

4 years

ERP/DWH Migration & Finance Processes

Deutsche Telekom

Analyzed impacts of the One.ERP Integration on DWH-T systems; mapped data flows across SAP and BW landscapes
Assessed core finance processes: revenue processing, month-end close, dunning/receivables, logistics, sub-ledgers, billing/invoicing
Prepared knowledge materials on controlling, treasury, accounting, financial reporting differences under IFRS/HGB; clarified KPIs (OCF, ROCE, CapEx, EBITDA)

Jan 2008 - Dec 2011

4 years

Banking & Finance Consulting

Multiple Institutions

Senior consultant for group-level management reporting: high-level and functional concepts, requirements alignment, project management
Advisory on reporting and BI flows supporting controlling and treasury needs; OSPlus migrations and data integration
Designed cash-flow prototypes and advised on cash management solutions and processes
Conducted business analysis and prototype design for closing workflows and financial reporting
Developed business plan, liquidity planning, and contribution margin calculation for an e-commerce startup; market model based on exchange-style mechanics

Jan 2008 - Dec 2009

2 years

BI/DWH Architecture & Reporting

NordLB

Consulted on BI/DWH architecture and reporting for banking functions with emphasis on controlling and risk-related reporting

Jan 2001 - Dec 2008

8 years

Consultant

Independent/Freelance

Managed projects across banking, media, and telecom; roles in project management, software engineering, and product management

Jan 1996 - Dec 2001

6 years

Co-founder and Manager

Scopas Medien AG / FunPage

Co-founded and managed agency operations; production planning; delivered multiple client projects
Conducted software and multimedia engineering

Industries Experience

See where this freelancer has spent most of their professional time. Longer bars indicate deeper hands-on experience, while shorter ones reflect targeted or project-based work.

Experienced in Media and Entertainment (27 years), Information Technology (14 years), Telecommunication (12 years), Banking and Finance (11 years), Automotive (5.5 years), and Transportation (3 years).

Media and Entertainment

Information Technology

Telecommunication

Banking and Finance

Automotive

Transportation

Business Areas Experience

The graph below provides a cumulative view of the freelancer's experience across multiple business areas, calculated from completed and active engagements. It highlights the areas where the freelancer has most frequently contributed to planning, execution, and delivery of business outcomes.

Experienced in Product Development (27 years), Information Technology (22 years), Business Intelligence (16 years), Research and Development (14 years), Project Management (11 years), and Operations (7 years).

Product Development

Information Technology

Business Intelligence

Research and Development

Project Management

Operations

Summary

Consultant and engineer with 20+ years in data engineering, AI/LLM integration, and cloud architectures. Specialized in scalable data pipelines, knowledge graphs, and retrieval-augmented generation (RAG). Deep background in NLP/LLM engineering, ontology integration, semantic enrichment, and graph algorithms. Track record across automotive, finance, telecom, public sector, and law enforcement.

Skills

Data Engineering: Aws (Glue, Athena, S3, Lambda, Redshift, Sagemaker, Amplify), Databricks, Ci/cd (Gitlab Ci/cd, Jenkins), Python, Pyspark
Ai & Llms: Llms (Openai, Mistral, Gemini, Claude, Amazon Nova), Rag Architectures, Embeddings, Vector Databases (Neo4j, Qdrant), Ontology Integration (Esco), Nlp (Spacy, Nltk), Generative Ai
Graph Analytics: Knowledge Graph Modeling, Graph Algorithms (Community Detection, Triangle Count, Louvain), Graph Reasoning, Semantic Enrichment, Graphql
Finance & Controlling Expertise: Ifrs/hgb Accounting, Controlling, Treasury, Cash Management, Risk Controlling, Financial Reporting

Languages

German

Advanced

English

Advanced

Education

Oct 1992 - Jun 1997

University of Applied Sciences Furtwangen

Diploma in Business Informatics (FH) · Business Informatics · Furtwangen, Germany

Oct 1988 - Jun 1991

Werner Heisenberg School

Abitur (German A-levels) · Rüsselsheim, Germany

Profile

Created

September 2025

Last Update

January 2026

Need a freelancer? Find your match in seconds.

Try FRATCH GPT

Frequently asked questions

Do you have questions? Here you can find further information.

Where is Stephan based?

Stephan is based in Wiesbaden, Germany.

What languages does Stephan speak?

Stephan speaks the following languages: German (Advanced), English (Advanced).

How many years of experience does Stephan have?

Stephan has at least 30 years of experience. During this time, Stephan has worked in at least 24 different roles and for 16 different companies. The average length of individual experience is 1 year and 3 months. Note that Stephan may not have shared all experience and actually has more experience.

What roles would Stephan be best suited for?

Based on recent experience, Stephan would be well-suited for roles such as: NLP/LLM Chatbot, AWS PoC / Content Analytics, Knowledge Graph / NLP / AI.

What is Stephan's latest experience?

Stephan's most recent position is NLP/LLM Chatbot at Insurance.

What companies has Stephan worked for in recent years?

In recent years, Stephan has worked for Insurance, Federal Criminal Police Office, Siemens, BMW Connected Vehicle, and Generative AI Blog & AWS Serverless Experiments.

Which industries is Stephan most experienced in?

Stephan is most experienced in industries like Media, Entertainment and Publishing, Information Technology (IT), and Telecommunication. Stephan also has some experience in Banking and Finance, Automotive, and Transportation and Logistics.

Which business areas is Stephan most experienced in?

Stephan is most experienced in business areas like Product Development, Information Technology (IT), and Business Intelligence. Stephan also has some experience in Research and Development (R&D), Project Management, and Operations.

Which industries has Stephan worked in recently?

Stephan has recently worked in industries like Information Technology (IT), Media, Entertainment and Publishing, and Government and Public Administration.

Which business areas has Stephan worked in recently?

Stephan has recently worked in business areas like Information Technology (IT), Product Development, and Research and Development (R&D).

What is Stephan's education?

Stephan attended University of Applied Sciences Furtwangen for Business Informatics.

What is the availability of Stephan?

Stephan is immediately available full-time for suitable projects.

What is the rate of Stephan?

Stephan's rate depends on the specific project requirements. Please use the Meet button on the profile to schedule a meeting and discuss the details.