Description

ML Systems Engineer | 8+ Years of End-to-End AI Production (LLMs, Vision, & Tabular)

Stop building ML experiments. Start deploying production systems.

Building a model in a notebook is only the first step. I bridge the gap between "experimental AI" and "production-grade software." Based in Germany, I help EU businesses design and maintain high-impact ML systems (LLMs, Vision, and Tabular) that deliver real-world ROI.

Multi-Modal Expertise Across the Pipeline

I specialize in building the right architecture for your specific data:

Generative AI & LLMs: Fine-tuning (QLoRA/LoRA), structured JSON extraction, and hybrid ML/Rule-based logic for deterministic results.
Computer Vision & Multi-Modal: Integrating visual features (CLIP/Qwen) into broader predictive pipelines.
Predictive Analytics: Designing multi-output regression systems using combined text, structured, and visual data.
The MLOps Foundation: I build the infrastructure that makes models work Airflow pipelines, BigQuery workflows, FastAPI backends, and CI/CD regression checks.

Why Work With Me?

Most specialists choose either Data Science or DevOps. I offer End-to-End Ownership:

Modeling Excellence: High-performance models (PyTorch, Hugging Face) focused on metrics that matter (RMSE, MAE, F1).
Operational Stability: I build the "shields" around models automated evaluation loops, and monitoring to prevent drift.

Technical Stack

AI: Python, PyTorch, Hugging Face, Scikit-learn, XGBoost.
Ops: Docker, Kubernetes, Airflow, BigQuery, FastAPI, CI/CD.

Proven Results

Hybrid Extraction: Combined LLMs and schema validation for 99%+ data accuracy.
Complex Regression: Built multi-feature pipelines (text+image+data) that significantly outperformed baselines.

Let’s Build a Trust-Based Partnership. I am ready to own your outcomes from design to deployment.

Industry field of expertise

Languages

English
Native or bilingual
German
Basic
Telugu
Native or bilingual

Workplace preferences

Can work on-site

Mannheim (up to 50km)

Capgemini Engineering Deutschland S.A.S & Co.KG
Consultant - MLOps Engineer | Data Scientist
DIGITAL AND IT
August 2020 - March 2025 (4 years and 7 months)
Munich, Germany
As a full-time Consultant - MLOps Engineer | Data Scientist, I worked across multiple client engagements in the Pharma and Aviation industries, delivering production-grade machine learning, MLOps, data, and backend solutions.

I was responsible for designing, building, and deploying end-to-end ML pipelines using Docker, Dkube, and Kubeflow, supporting model training, retraining, versioning, and deployment. I collaborated closely with data scientists and subject matter experts to gather requirements, refactor research notebooks into structured, maintainable codebases, and ensure smooth operationalization of ML models in production environments.
In addition to MLOps, I contributed to DevOps and automation initiatives, developing Jenkins pipelines for automated regression testing and integrating custom Python-based comparison tools into CI/CD workflows.
As part of data engineering efforts, I designed and automated data pipelines on Google Cloud Platform, leveraging BigQuery, Airflow, and LookML to process, transform, and onboard data for analytics and reporting platforms. This included setting up BigQuery views, LookML models, metadata structures, and automated reports to support business intelligence use cases.
I also developed backend services and applications using Python, FastAPI, and REST APIs, enabling data driven workflows and scientific applications. My work included building ML prototypes such as automatic license plate extraction systems using OCR, object detection, and transfer learning.
Overall, my role focused on bridging data science, engineering, and business needs, delivering scalable, reliable, and well documented solutions within enterprise environments.
MLOps DevOps & CI/CD Python (Programming Language) Docker FastAPI
CAMELOT ITLab GmbH
Consultant - Associate Data Scientist
DIGITAL AND IT
April 2018 - July 2020 (2 years and 3 months)
Mannheim, Germany
As a full-time AssociateData Science Consultant, I worked on multiple analytics and machine learning engagements within the Chemicals industry, focusing on time series forecasting, deep learning, e-commerce analytics, and document automation.
I developed time series forecasting solutions on multivariate data, managing end-to-end workflows including data preprocessing, feature engineering, and external data causality and correlation analysis. I trained and evaluated models such as linear regression, decision tree regressors, XGBoost, and vector autoregression (VAR) to support demand and trend forecasting use cases.
In a related commitment, I built a deep learning demand pattern classification model using TensorFlow and convolutional neural networks (CNNs) to address seasonal demand variability and reduce safety stock costs. The model achieved a validation accuracy of 84.77% and was presented at SAP Sapphire 2018, demonstrating both technical impact and business value.
For an e-commerce analytics project, I developed sales and returns analytics, conducted A/B testing, and built text classification models to detect duplicate and similar products using titles and descriptions. I handled data collection, preprocessing, model training, tuning, and deployment using Python, SAP HANA, and SAP Analytics Cloud, and designed automated ETL pipelines to deliver reliable datasets and KPI dashboards with model evaluation and drift monitoring.
Additionally, I delivered an automated invoice data extraction solution, building an end-to-end pipeline for image ingestion, preprocessing (deskewing and denoising), OCR execution, and structured key-field extraction into standardized JSON outputs. I fine-tuned models using transfer learning, tracked performance metrics, and conducted detailed error analysis to improve accuracy.
Overall, my role focused on delivering practical, production-ready data science solutions aligned with real business outcomes.
OCR Text Mining & Deep Learning DevOps & CI/CD Python (Programming Language) SAP Analytics Cloud
SAP Deutschland SE & Co. KG
Working Student
DIGITAL AND IT
February 2016 - July 2017 (1 year and 5 months)
Walldorf, Germany
As a working student, I contributed to internal capability building by organizing and delivering a beginner-level data science training program. I designed and curated structured learning materials, including slides, notebooks, and datasets, and conducted live training sessions focused on practical machine learning concepts and hands-on exercises.
I supported participants by providing solutions, guidance, and troubleshooting for machine learning questions and lab assignments, ensuring strong conceptual understanding and applied learning. In parallel, I developed Python-based automation scripts to streamline daily administrative and operational tasks related to the data science course and the Development Expert curriculum, significantly reducing manual effort.
I also collected and analyzed participant feedback and tracked learning progress to continuously refine the curriculum, exercises, and supporting automation workflows. This initiative helped improve learning outcomes while establishing a more scalable and efficient internal training framework.
Python (Programming Language)