Hi, I'm Tomislav Maruscak

Expert Data Scientist & AI Engineer

10+ years of experience delivering production-grade ML, forecasting, and optimization solutions. Proven leader achieving step-change accuracy improvements and multi-million-euro cost savings.

10+
Years Experience
€MM
Cost Savings Delivered
7
Industries Served

About Me

I'm an expert Data Scientist and AI Engineer with 10+ years of experience delivering production-grade machine learning, forecasting, and optimization solutions in global enterprise environments.

Currently based in Berlin, I specialize in time series analysis, demand forecasting, logistics and inventory optimization, and SAP-centric data landscapes. My work has delivered measurable business impact, including multi-million-euro cost savings and step-change forecast accuracy improvements (e.g., MAPE reduced from ~0.7 to ~0.13).

I combine deep mathematical rigor with pragmatic AI engineering to deliver robust, explainable, and adoptable solutions. As a certified Business Information Security Officer, I ensure all solutions meet enterprise security and compliance standards.

Fluent in Croatian (native), German (C1), English (C1), and Spanish (C1), I work effectively with international teams across diverse markets.

ISO/IEC 27001 Security Officer
GCP Associate Cloud Engineer
MSc Statistics
Tomislav Maruscak - Expert Data Scientist & AI Engineer

Core Competencies

Data Science & AI

  • Time Series Forecasting & Optimization
  • Foundation Models (TimesFM, Chronos)
  • Deep Learning (DeepAR, TensorFlow)
  • Predictive & Probabilistic Modeling
  • MLOps & Autonomous Retraining
  • Explainable & Trustworthy AI

Data Engineering

  • Scalable ETL/ELT Pipelines
  • Data Warehousing & Architecture
  • Big Data & Streaming Processing
  • Azure Synapse, BigQuery, PostgreSQL
  • SAP Hana, SAP BTP & Kyma

Cloud & DevOps

  • Google Cloud Platform (GCP)
  • Microsoft Azure (AzureML)
  • Amazon Web Services (AWS)
  • Docker & Kubernetes
  • Git, CI/CD

Programming

  • Python (Polars, Pandas, Scikit-learn)
  • R (Statistical Computing & Shiny)
  • SQL & SQLAlchemy (Advanced)
  • C/C++
  • API Development & OAuth 2.0

Specialized Tools

  • TimesFM, Chronos, DeepAR, Prophet
  • Optuna, MLflow, NetworkX
  • GPT-5.1 & LLM Integration
  • Dash, Plotly, R Shiny
  • Docker, Git, CI/CD Tools

Industry Expertise

  • Aviation & Air Freight
  • MRO Logistics & Optimization
  • Smart Cities & IoT
  • E-commerce & Sales Analytics
  • Waste Management & Sustainability
  • Energy & Industrial Manufacturing

Professional Experience

Delivering measurable business impact through production-grade AI solutions

01/2026 - 05/2026

Data Scientist / AI Ops / Project Lead

Lufthansa Technik - MRO Logistics

Forecasted repair duration of aircraft components to optimize planning and inventory management.

  • Achieved step-change accuracy improvement: Reduced MAPE from ~0.7 to 0.13
  • Enabled multi-million-euro inventory cost savings by correcting systematic over-forecasting
  • Designed scalable ETL pipelines for high-quality time series datasets
  • Benchmarked classical statistical models vs. foundation models (TimesFM, Chronos)
  • Delivered fully autonomous retraining pipeline with explainability dashboard
Polars Chronos TimesFM ARIMA AzureML Synapse
07/2025 - 01/2026

Data Scientist / Project Lead

Lufthansa Technik - MRO Logistics

Optimized warehouse stock replenishment strategies based on US tariffs and dynamic cost components.

  • Translated complex business challenge into rigorous mathematical formulation
  • Introduced dynamic cost factors reflecting volatile global market conditions
  • Developed multi-objective optimization framework for cost-optimal routing
  • Achieved substantial performance gains through early elimination of suboptimal solutions
Polars NetworkX Optimization AzureML SAP
01/2025 - 05/2025

Data Scientist / AI Engineer

Lufthansa Technik - Waste Management

Dynamic AI-based document processing for cost distribution across waste management operations.

  • Consolidated invoices from varying waste processing suppliers
  • Evaluated different AI models for intelligent invoice parsing
  • Classified line items and established cost allocation to cost centers
  • Leveraged GPT-5.1 for advanced document understanding
GPT-5.1 Polars AI Parsing AzureML
07/2024 - 12/2024

Data Scientist / AI Ops / Project Lead

Lufthansa Technik - MRO Logistics

Forecasted exchanges in contractual and non-contractual settings to optimize logistics planning.

  • Consolidated diverse data sources and designed workarounds for data quality issues
  • Established data pipelines and AI Ops flows on streaming data
  • Implemented scalable production-ready solutions with explainability dashboard
  • Led team of consultants ensuring timely delivery of milestones
Polars TimesFM DeepAR Chronos AzureML Trustworthy AI
01/2024 - 08/2024

Data Scientist / ML Engineer / Project Lead

Lufthansa Technik - MRO Operations

Improved predictability of delivery times and routings to support on-time service delivery.

  • Managed team and defined KPIs using statistical tools to recognize data patterns
  • Implemented model robustness against minor changes while capturing data drift
  • Built Big Data pipelines with MLOps and CI/CD integration
Scikit-learn Polars Dash Azure Synapse Clustering
11/2023 - 03/2024

Data Scientist / ML Engineer / Project Lead

Lufthansa Technik - Market Intelligence

Estimated market availability of MRO parts considering supplier reliability for dynamic warehouse planning.

  • Facilitated workshops to identify use cases with customers
  • Automated tedious manual processes into streamlined data products
  • Applied probabilistic modeling and clustering techniques
Polars SQL AzureML Probabilistic Modeling
09/2023 - 12/2023

Data Scientist / ML Engineer

Lufthansa Cargo - Air Freight

Enhanced ML model performance and built user-friendly GUI for data product.

  • Redesigned ETL process with direct database access for on-demand data loading
  • Incorporated explainable AI elements for better model interpretation
  • Implemented user authentication system and model versioning with MLflow
Scikit-learn Optuna MLflow Dash OAuth2
11/2022 - 05/2023

Data Engineer

Tchibo - Food & Beverage

Redesigned existing data mart following star-schema best practices for multi-country sales data.

  • Reconstructed SAP Hana logic in GCP BigQuery
  • Optimized existing ETL jobs by consolidating interfaces
  • Ensured business continuity for critical data science use cases
GCP BigQuery SAP Hana SQLAlchemy Docker
2017 - 2022

Data Scientist

Siemens AG - E-Commerce & Smart City

Led end-to-end data science initiatives across sales analytics, customer intelligence, and IoT solutions.

  • Owned and delivered multiple data-driven decision support systems from PoC to production
  • Developed churn prediction, customer similarity, and recommender systems
  • Built demand forecasts and dynamic pricing models for e-commerce
  • Created predictive health monitoring for IoT devices using sensor data
  • Mentored junior colleagues and established data science best practices
AWS R Shiny Prophet SAP Hana IoT
2020 - 2021

Lecturer (Part-Time)

IMC University of Applied Sciences Krems

Taught graduate-level courses in data science and analytics.

  • Advanced Statistical Methods for Data Science
  • Data Visualization
  • Capstone Project Supervision
Teaching Statistics Data Science
2016 - 2017

Data Science Intern

Lufthansa Technik

Developed predictive health monitoring PoC for UAV components using sensor and time series data.

Python Sensor Data Predictive Modeling

What Colleagues Say

"

"Tomislav's ability to achieve step-change forecast accuracy improvements while delivering multi-million-euro cost savings demonstrates his exceptional combination of technical excellence and business acumen. His solutions are always production-ready and maintainable."

Senior Manager
Lufthansa Industry Solutions
"

"His ability to translate highly complex business problems into rigorous mathematical formulations is unmatched. Tomislav leads workshops that align stakeholders and decompose challenges into executable workstreams with precision."

Project Director
Lufthansa Technik
"

"What sets Tomislav apart is his deep mathematical rigor combined with pragmatic engineering. He delivers robust, explainable AI solutions that business stakeholders trust and adopt. His work directly impacts the bottom line."

Lead Data Scientist
Siemens AG
"

"Tomislav's expertise in MLOps and autonomous retraining pipelines transformed our team's capabilities. His mentoring established best practices that continue to drive quality and efficiency across all our data products."

Engineering Manager
Lufthansa Industry Solutions

Let's Work Together

Looking for a freelance data scientist or AI engineer for your next project? Let's discuss how I can help you achieve your goals.

Location

Berlin, Germany (Remote Available)

Languages

Croatian, German, English, Spanish

This helps us prevent spam submissions