About
I'm a freelance Cloud Engineer passionate about building cutting-edge cloud infrastructure and software solutions. Over the last decade, I've had the privilege of working across a diverse spectrum of environments — from start-ups and small companies to consultancy agencies and global corporations These experiences have honed my ability to adapt, innovate, and deliver results in dynamic settings.
Beyond client work, I share my expertise with the developer community by writing clear and practical technical articles on Medium. I also work on my own SaaS project, which is gradually coming to life
When I’m not coding or designing infrastructure, you’ll likely find me on a trail hike, pushing my limits on long-distance runs, or tinkering with my 3D printer to bring creative gadgets to life.
Experience
I am as a self-employed IT consultant and community leader, helping companies build and maintain robust, scalable and cost-efficient data driven systems in the cloud.
- Cloud Migrations
- IT Consulting
- Infrastructure as Code
- Cost Optimization
- DevOps
Developed an in-house ML based audio transcription service leveraging state-of-the-art AI models. Deployed the workload using AWS SageMaker with a custom containerized setup. Later scaled it with Kubernetes and GPU virtualization for better cost and performance. Achieved 10x cost reduction and 3x processing speed improvement, boosting client acquisition and retention. Lead an infrastructure team to extend company processes and cloud infrastructure for SOC2 Type II compliance.
- AWS
- Terraform
- Kubernetes
- Docker
- Python
I worked on improving and expanding an internal MLOps platform, enabling faster experimentation and efficient model deployment. Collaborated with the PO and senior DevSecOps engineers to extend cloud infrastructure in a compliant manner. Played a key role in scaling the platform, developing and deploying a scalable codebase and infrastructure for automatic provisioning and resource management for onboarded ML teams. Contributed to the hiring process by conducting technical interviews for the expansion of the MLOps team.
- Azure EntraID
- Azure DataBricks
- Azure Virtual Network
- Azure Virutal Machines
- Kubernetes
- Docker
- Python
- Jenkins
- Spark
I was part of the Data & AI team where I worked on a showcase project presenting the capabilities of end-to-end ML systems in the industrial analytics domain. I created a ML model for time-series classification of IoT sensor vibration data, developing a continuous-learning ML system and deployed the setup to production leveraging serverless architecture for minial infrastructure cost.
- AWS Lambda
- AWS Kinesis
- AWS DynamoDB
- AWS GreenGrass
- PyTorch
- Python
- Jupyter
Conducted advanced research on Cyber Security and Large Language Optimization topics. Developed a novel method to identify and prune redundant model components post-training, reducing complexity while maintaining performance. Also contributed to the development of a state-aware security vulnerability scanner using AI and ML techniques.
- Scientific Research
- Technical Writing
- Linux
- OWASP Top 10
- Machine Learning
- Data Science
I led a venture in collaboration with an industry partner to address a domain-specific challenge for industry 4.0. Oversaw all technical aspects for the development of an innovative IoT device and a web application for monitoring: Hardware Development: Designed and built an ESP-32-based IoT environmental sensor for automated daily data transmission to a remote server. Web Application Development: Developed a React-based web application to provide users with analysis and insights from the collected IoT data, hosted in an AWS environment. Infrastructure Management: Set up and managed cloud infrastructure for secure data storage and IoT device networking.
- AWS EC2
- AWS Route 53
- AWS VPC
- React
- MongoDB
- PostgreSQL
- Python
- Typescript
- NodeJs
Projects
DevLocus: CloudOps, DevOps, and MLOps topics
My comunity platform for discussing the latest trends, tools, and best practices in CloudOps, DevOps, and MLOps with the main goal of creating collaboration and sharing knowledge among professionals in the field.
SWaTEval: An Evaluation Framework for Stateful Web Application Testing
A research project from my contribution at Fraunhofer IOSB, SWaTEval uses ML-based fuzzy state extraction to uncover vulnerabilities in web applications by analyzing complex state-dependent behaviors.
CoffAI: Coffee Type Classification from Vibrational Data
The showcase project for Duesentrieb Lab at esentri where I was contributed with the model package and cloud deployment. Credits go to esentri AG for the project ownership.
Differentiable Slimming for Memory-Efficient Transformers
A research project from my contribution at KIT CES where I was working on a novel pruning technique for LLMs from the GPT and BERT family to reduce memory consumption while maintaining performance.
