08.01.2026 aktualisiert

**** ******** ****
verifiziert
Premiumkunde
100 % verfügbar

Ph.D. Data Scientist & Rust Engineer | High-Performance Forecasting (500k+ SKUs)

Biberach, Deutschland
Deutschland +2
info: Deutschland, Österreich, Schweiz
Dr. rer.nat. Mathematik
Biberach, Deutschland
Deutschland +2
info: Deutschland, Österreich, Schweiz
Dr. rer.nat. Mathematik

Profilanlagen

Lebenslauf - Simon Müller.docx
Technical Portfolio_ High-Performance AI & Forecasting.pdf

Über mich

Ph.D. Statistician & Rust Engineer. I combine rigorous statistics (CMC, Bayesian) with high-performance software (Rust, DuckDB, MCP). specialized in scalable forecasting (Chronos-2) & AI tooling. 12+ years of experience delivering production systems for Kärcher, Hensoldt & Boehringer.

Skills

Agile MethodologieKünstliche IntelligenzAmazon Web ServicesAmazon Elastic Compute CloudAmazon S3Data AnalysisApache HTTP ServerSystems EngineeringArchitekturMicrosoft AzureBigqueryC++Cloud ComputingClusteranalyseContinuous IntegrationETLData WarehousingNachfrage-PrognoseDevopsIngenieurwesenExperimentierenForecastingGithubR (Programmiersprache)PythonMachine LearningObjektorientierte Software-EntwicklungVisualisierungPower BiAzure Data LakeSAP ApplicationsSQLStatistikenZeitreihenanalyseRust (Programming Language)Jupyter NotebookPytorchGenerative AIAws LambdaGitPandasScikit-learnUnternehmens-IntegrationMachine Learning OperationsDocker
CORE COMPETENCIES
High-Performance Computing & Systems Engineering: Specialised in migrating slow Python pipelines to ultra-fast architectures.
Languages: Rust (Crates.io contributor), C++, Python (Polars/Pandas), SQL.
Tech: DuckDB (Extension Development), WebAssembly (WASM), Apache Arrow.
AI Engineering & GenAI: Building production-grade AI tools and inference engines.
Generative AI: Chronos-2 (Rust Port), Model Context Protocol (MCP) Server development, AI Agents (Beads).
Frameworks: PyTorch, Scikit-learn, Stan.
Advanced Statistics (PhD):
Methodologies: Bayesian Statistics, Functional Data Analysis (FDA), Regression, Machine Learning, Design of Experiments (DoE).
Focus: Large-scale Time Series Forecasting (Hierarchical/Sparse), Uncertainty Quantification.
Cloud & DevOps:
AWS: SageMaker, S3, ECR, Lambda.
Infrastructure: Docker, GitHub Actions (CI/CD), Azure Data Lake.
Enterprise Integration:
Systems: SAP Ecosystem (RFC/ODP), Kinaxis RapidResponse, MS Dynamics 365.
Website

Sprachen

DeutschMutterspracheEnglischverhandlungssicher

Projekthistorie

CMC Statistics & Spectral Data Analysis (Pharma R&D)

Pharma und Medizintechnik

>10.000 Mitarbeiter

Development of novel statistical methods for analysing complex spectral data using Functional Data Analysis (FDA) and Partial Least Squares (PLS). Enhancing Quality Control (QC) strategies and modelling for GMP environments.

Automated Financial Forecasting (Hierarchical ML & Solara UI)

Industrie und Maschinenbau

1000-5000 Mitarbeiter

Replaced manual Excel workflows with an automated, unbiased ML pipeline for Order Intake, Revenue, and Cash Flow. Built a custom Self-Service UI (Solara) enabling finance teams to run simulations independently. Stack: Python, Solara, Hierarchical Forecasting, Causal Models.

Automated Data Quality & Anomaly Detection Pipeline

Industrie und Maschinenbau

>10.000 Mitarbeiter

Automated reporting system to detect data anomalies early. Increases confidence in downstream ML forecasts and reduces the risk of costly supply chain errors.
Stack: Python, AWS S3, Quarto, Automated Reporting.

ERP Data Engineering & Supply Chain Analytics (MS Dynamics 365)

Industrie und Maschinenbau

250-500 Mitarbeiter

Engineered a centralised data architecture surrounding the introduction of Microsoft Dynamics 365.

Key Deliverables:
  1. Data Unification: Architected a "Single Source of Truth" by extracting and unifying complex ERP data streams. This enabled the purchasing department to run automated risk analyses (stockouts vs. excess inventory).
  2. Cloud Integration: Built robust data preparation workflows for the migration to Azure, feeding forecasts directly back into the ERP ecosystem.
  3. Impact: Reduced manual workload and operational costs by automating demand planning processes and establishing a scalable foundation for future analytics.
Stack: Python, SQL, MS SQL Server, MS Dynamics 365, Azure Cloud, Data Engineering, ETL.

Spare Part Intelligence Engine (High-Performance ML)

Industrie und Maschinenbau

1000-5000 Mitarbeiter

Development of a specialised forecasting engine for spare parts to minimise stockouts and improve equipment uptime. Focus on handling intermittent demand patterns and optimising safety stock levels.
Stack: Python, AWS, Statistical Modelling.

High-Scale Demand Forecasting (500k+ SKUs)

Konsumgüter und Handel

1000-5000 Mitarbeiter

Architecture and deployment of an end-to-end ML pipeline on AWS. Processing 500,000+ products with results integrated directly into Kinaxis RapidResponse. Reduced inventory holding costs and optimised global supply chain planning.
Stack: Python, AWS SageMaker, Kinaxis Integration, Big Data.

Global Supply Chain Automation (Dockerized Pipeline)

Industrie und Maschinenbau

5000-10.000 Mitarbeiter

Implementation of a global forecasting framework for ~500k Material Numbers. The pipeline is fully containerised and managed via AWS SageMaker, requiring minimal maintenance from the client IT team. Stack: Python, Docker, AWS SageMaker, CI/CD.

High-Performance Forecasting Engine (Rust & WASM)

Industrie und Maschinenbau

>10.000 Mitarbeiter

Engineered a client-side forecasting tool using Rust compiled to WebAssembly (WASM). Allows complex revenue forecasting directly in the browser (Google Workspace) without heavy backend infrastructure.
Stack: Rust, WebAssembly (WASM), GitHub Actions.

Principal CMC Statistician & Quality Strategy Lead

Boehringer Ingelheim & Co KG

Pharma und Medizintechnik

Served as the Principal Statistician for CMC (Chemistry, Manufacturing, and Controls), driving statistical strategy in a strictly regulated GMP environment.

Key Deliverables:
  1. Strategic Leadership: Developed and implemented statistical data analysis strategies for BioPharma production, including training Subject Matter Experts (SMEs) in statistical decision-making.
  2. Process Optimisation: Applied Design of Experiments (DoE) and advanced sampling models to identify sources of variability, ensuring robust process validation and equivalence testing.
  3. Advanced Methodologies: Introduced Bayesian statistics to enhance decision-making in complex quality scenarios.
  4. Quality Assurance: Conducted method validation and root cause analysis via advanced data visualisation to demonstrate process performance and ensure compliance.
Stack: CMC (Chemistry Manufacturing and Controls), GMP, Bayesian Statistics, Design of Experiments (DoE), Process Validation, Biostatistics, Root Cause Analysis, DesignExpert, R.

Automated Demand Forecasting & Reporting Pipeline (Glass Manufacturing)

Schott AG

Industrie und Maschinenbau

>10.000 Mitarbeiter

Developed a predictive analytics solution to anticipate customer order timing for glass tube products, enabling better alignment between Sales and Production.

Key Deliverables:
  1. Time Series Forecasting: implemented statistical models in R to predict specific order windows, allowing sales teams to optimise weekly planning.
  2. Automated Reporting: Leveraged Quarto to generate and distribute high-quality, automated forecast reports to stakeholders, ensuring consistent communication of production needs.
  3. Containerization: The entire pipeline (Data Prep → Modelling → Quarto Reporting) was dockerized for reliable execution in the Azure cloud.
  4. Feature Engineering: Integrated sales territory data and customer interaction history to refine model accuracy.
Stack: R, Quarto, Time Series Forecasting, Docker, Azure Cloud, Automated Reporting, Supply Chain Analytics.

High-Scale Smart Meter Forecasting & Anomaly Detection (Databricks)

E.ON Business Services GmbH

Energie, Wasser und Umwelt

>10.000 Mitarbeiter

Architected a scalable analytics platform on Azure Databricks to process multi-year, high-frequency data from over 10,000 smart meter devices.

Key Deliverables:
  1. Big Data Pipeline: Implemented a distributed processing pipeline using Apache Spark to handle massive time-series data ingest and transformation.
  2. Model Optimisation: Conducted extensive benchmarking of Statistical methods (ARIMA, ETS), Machine Learning (RandomForest, XGBoost), and Deep Learning (Keras) to determine the optimal production architecture.
  3. Production Deployment: Deployed a robust solution for forecasting and real-time anomaly detection, enabling proactive issue resolution and optimised energy management.
Stack: Apache Spark, Databricks, R, Deep Learning (Keras), Time Series Forecasting, Anomaly Detection, Big Data.

Solar Asset Predictive Maintenance & Anomaly Detection (R/Shiny)

E.ON Business Services GmbH

Energie, Wasser und Umwelt

>10.000 Mitarbeiter

Engineered a Proof-of-Concept (PoC) system to detect underperforming B2C solar assets by distinguishing between environmental factors and technical faults.

Key Deliverables:
  1. Hybrid Modelling: Combined Physical Models of solar generation with statistical Time Series forecasting to accurately identify anomalies in strongly correlated data.
  2. External Data Fusion: Integrated high-resolution weather data (irradiance, wind speed, temperature) to normalise performance metrics and reduce false positives.
  3. Interactive Diagnostics: Built an R Shiny dashboard allowing engineers to visually explore asset performance, validate anomalies, and prioritise maintenance schedules.
  4. Impact: Validated a strategy to maximise ROI on solar infrastructure by enabling proactive rather than reactive maintenance.
Stack: R, R Shiny, Predictive Maintenance, Anomaly Detection, Physical Modelling, Time Series Analysis, IoT Analytics.

AI-Driven Sales & Production Forecasting Engine (Azure)

Energie, Wasser und Umwelt

50-250 Mitarbeiter

Architected and deployed an end-to-end predictive analytics pipeline on MS Azure to forecast customer order timing.

Key Deliverables:
  1. Predictive Engine: Developed models to predict specific "Next Order Dates" for customers, enabling sales teams to proactively plan weekly schedules and target high-probability accounts.
  2. Operational Synchronization: The system aligns manufacturing processes and inventory levels with predicted demand, reducing operational bottlenecks.
  3. Cloud Automation: The entire pipeline (Data Prep → Modeling → Delivery) is containerized via Docker and runs on fully automated weekly (Sales) and monthly (Revenue/Production) schedules.
  4. Advanced Modeling: employed extensive feature engineering (customer interactions, territorial segmentation) to maximize accuracy.
Stack: Python, Docker, MS Azure, Predictive Analytics, Sales Forecasting, Feature Engineering, Data Engineering.

Advanced Analytics & Data Science Consulting

Daimler Financial Services

Banken und Finanzdienstleistungen

1000-5000 Mitarbeiter

As part of the core team establishing Data Science at DFS, I acted as a technical lead and internal consultant, delivering complex feasibility studies and upskilling the department.

Key Deliverables:
  1. High-Performance Computing: Taught and implemented C++ integration into R to accelerate heavy analytical workloads, establishing early standards for reproducible and high-speed code (R Packages).
  2. Financial Modelling: Developed advanced statistical models for Credit Risk (Insolvency detection via Ensembling/Neural Networks) and Insurance Pricing (Tobit models/Targeted Learning for safety equipment impact).
  3. Agile Transformation: Co-architected the team's move to SCRUM, implementing Jira/Confluence to streamline collaboration for the Data Science unit.
  4. Diverse Use Cases: Delivered production concepts for Fraud Detection, Customer Segmentation (Graph Algorithms), and Record Linkage on MS Azure.
Stack: R, C++, Microsoft Azure, Credit Risk Modelling, Fraud Detection, Machine Learning (XGBoost, SVM), Agile (SCRUM), Financial Services.

High-Performance Statistical Anomaly Detection (C++ / R Package)

Daimler Financial Services

Banken und Finanzdienstleistungen

1000-5000 Mitarbeiter

Engineered a custom R package aimed at automated data quality control for a global contract management system.

Key Deliverables:
  1. Performance Engineering: Integrated C++ algorithms into R to handle high-dimensional financial data, overcoming the performance bottlenecks of standard statistical libraries.
  2. Unsupervised Learning: Implemented ensemble-based subspace search methods to detect hidden "unknown unknown" outliers that rule-based systems missed.
  3. Software Standards: Delivered a modular, reusable library complete with CI/CD pipelines (GitLab), ensuring long-term maintainability for the client's analytics team.
Stack: R, C++, Statistical Learning, Anomaly Detection, Data Quality, Financial Services, CI/CD, Unsupervised Learning.

Zertifikate

AWS Certified Cloud Practitioner

Amazon Web Services Training and Certification

2023

Kontaktanfrage

Einloggen & anfragen.

Das Kontaktformular ist nur für eingeloggte Nutzer verfügbar.

RegistrierenAnmelden