%20copy.jpg)
ABOUT
AI SOLUTION ARCHITECT SENIOR ML ENGINEER
I have a PhD in Natural Language Processing with 9-10 years of applied hands-on experience in AI and Machine Learning. I have developed or led the development of robust AI solutions for enterprises such as Databricks, Siemens, as well as numerous SMEs and dynamic startups.
SKILLS & EXPERTISE
Areas of Expertise
-
AI Solution Architecture
-
AI agents and agentic architectures. Agent orchestration, guardrails, and evaluation
-
Large Language Models (LLMs): Building and fine-tuning
-
Natural Language Processing (NLP)
-
Evaluation methods. LLM-as-a-judge. LLM jury
-
Data quality. Agentic data annotation and dataset creation pipelines.
-
Distributional Semantics, Word Embeddings, and Retrieval Augmented Generation (RAG)
-
Traditional machine learning and statistical models
Technical Skills
-
Core AI & ML Engineering — Expert
-
Python • PyTorch • HuggingFace • Anthropic API • OpenAI Agents SDK • Fine-tuning (instruction / preference) • Hyperparameter optimization •
Vector DBs • RAG • Weights & Biases • FastAPI • Pydantic -
Cloud & Infrastructure — Expert AWS • Hetzner Cloud • OpenAI and Anthropic APIs • GCP • LLMOps • MLOps • Docker • Microservice architectures • CI/CD • Ansible • GitHub
Sep 2024 – Present
ML Lead & Architect
Calibrion AI - Berlin DE
Designed the core architecture of Calibrion, the first agent-native data annotation platform specialized for LLMs.
Led end-to-end product development from backend (Django, Python, FastAPI, PostgreSQL) and infrastructure (Docker, CI/CD, Ansible) to frontend (React and Javascript).
Developed and integrated AI/ML and data science components, including AI agents, evaluation metrics, annotation workflows, and quality-control pipelines.
Drove strategy and execution as founder, managing freelancers, defining product roadmap, communicating with target audience, negotiating deals, and shaping positioning in the emerging LLM ecosystem.
August 2021 – Present
AI Architect | Senior ML Engineer /
Freelance
Senior ML engineer | AI architect
Simplica Corp · Wilmington · Remote
Oct 2024 - PresentImplemented fine-tuning and experimentation infrastructure for fine-tuning various LLMs for generating framework-specific code.
Designed and implemented an agentic sub-system for interacting with web components and integrated it into their existing framework.
Skills: Python, LLMs, fine-tuning (instruction-tuning, preference tuning), LLM data, synthetic data, data quality, OpenAI's Agents SDK, AWS Bedrock, Weights & Biases, Agentic architectures. Multi-agent systems.
Senior ML engineer
Crossfader Inc · London · Remote
June 2025 - August 2025Implemented a RAG system capable of chatting with their uses about their course contents form 1000+ video course transcripts in 11 languages.
Optimized the retrieval through various vector adjustments, re-ranking, similarity metrics, and Qdrant nested filters.
Skills: Python, Qdrant, Distributional Semantics, AWS, Docker, Streamlit.Senior ML and data science consultant
Databricks · San Francisco · Remote
May 2024 - August 2024Analyzed the accuracy and reliability of annotations for a specific use case which was
foundational to informed decision-making within the business.Provided the client with data-
driven advice on the status of their data and proposed approaches to improve the quality of
their data collection pipeline.Created a customized roadmap for their ML experimentation.
Skills: Evaluation methods, Data annotation, Classification, Active learning, Online learning.
Senior ML engineer for LLM customization and integration
Atreus GmbH · Munich, Germany · Remote
Nov 2022 - Dec 2023Developedmultiple LLM-based POCs for their CV and interview processing use cases.
Defined their AI development roadmap focusing on LLMs and high-stake business use cases.
Improved their existing ML code repos by implementing CI/CD, unittests, code quality
checks, and SOLID design principles.
Skills: Python, Pytest, GitHub actions, Airbyte, LLMs, AWS, ELT, bloom-560m, gpt-neo, Streamlit,
GPT-3.5, Swagger, Docker.
Senior ML engineer in NLP and LLM
Schultz Family Foundation · Seattle, US · Remote
Aug 2021 - Jun 2022Developed a question-answering conversational model with access to high quality content to
supports entrepreneurs in their business journey.Developed various ML components for the
mentioned system, including classifiers, key phrase extractors, and clustering models.
Skills: LLMs, BERT, T5, GPT-3, Flask, Fastapi, GCP, Azure, Question Answering
(MRC), Dialogue Systems, Python, PyTorch, Plotly, scikit-learn.
Oct 2021 – Dec 2022
ML Lead /
Archipelo - SF, USA - Remote
Defined ML goals in close collaboration with the C-suite. Translated business needs to ML
features with product management.Led the development of ML and NLP infrastructure as a hands-on and collaborative team lead. Developed software-library-name-highlighting (a key feature of the product) that exceeded the required 75% accuracy target by 5%.
Skills: AI strategy, Python, PyTorch, CNNs, scikit-learn, Weights & Biases, GCP, Docker, Poetry,Data Annotation. Classification, NER, Collaborative and Hands-on Leadership.
Oct 2018 – JUNE 2020
Data Scientist /
OMNIUS - BERLIN, DE
Improved the accuracy of their document classifier by 12%. Implemented transfer learning and
anomaly detection for doc/text classification and sequence tagging in order to improve their
performance.Carried out statistical analysis of data for improving the performance of
downstream ML models.Co-advised MSc theses on Named Entity Recognition and active learning. Developed a Siamese text-to-text mapping service.
Skills: Python, PyTorch, TensorFlow, NER, text classification, statistical NLP, anomaly detection, document processing, data annotation,annotation evaluation (IAA).
Feb 2017 – Sep 2018
Data Scientist /
MARKET LOGIC SOFTWARE - BERLIN, DE
Developed a natural language understanding, generation, and a reinforcement-learning-based dialogue manager for a chatbot that answers marketing questions based on market research data. This research-heavy project was developed for and presented to Unilever ass a pilot.
Skills: Python, Java, Kafka, MongoDB, reinforcement learning, sequence-to-sequence models, attention mechanism, AWS
WORK EXPERIENCE
EDUCATION
PhD in Computer Science
University of Geneva, Switzerland / Graduated:2017
Specialization in natural language processing
with highest distinction
Master of Science in Computer Sceince
University of Lugano, Switzerland / Graduated:2010
CERTIFICATIONS
Certified ScrumMaster (CSM)
AWS Certified Solutions Architect– Associate
Oracle Certified Professional,
Java SE 8 Programmer
Twelve Angry LLMs
An extensible Python library for creating and using LLMs as judges --an emerging topic in generative AI where LLMs are used to evaluate other LLMs or AI agents. This project providing a framework for defining adifferent types of judges, from those that return a simple score to those that provide a detailed, descriptive evaluation.
Wordview
Wordview is an open-srouce Python package for Exploratory Data Analysis of text and provides many statistics about your data in the form of plots, tables, and descriptions allowing you to have both a high-level and detailed overview of your data. Recently, LLMs were integrated so that you can now chat with your text corpus' statistics.