Hi, I'm Sebastian — an AI Data Platform Engineer who makes real-world AI actually work. That means I don't just prompt a model and hope for the best — I design and build full-stack systems that augment human workflows and automate repetitive work, reliably, cleanly, and at scale.
As an AI Data Platform Engineer, I operate at the intersection of data infrastructure, machine learning, and operational AI — turning raw data and ideas into intelligent systems that are scalable, maintainable, and production-ready. My background runs deep: I started in Data Science, moved into Machine Learning Engineering, and built solid MLOps foundations long before prompt engineering became a LinkedIn headline.
I've shipped AI agents that review products 10x cheaper and 100x faster than humans. I've built tools that help retailers plan categories 12x faster and 5x cheaper. I've led the AI behind enrichment platforms that take fragmented product data and turn it into clean, actionable insight — using LLMs, fuzzy matching, and human-in-the-loop systems that actually scale.
What sets me apart? I explain complex ideas simply, ship working systems fast, and stay laser-focused on business value. I've worked across industries, always bringing fresh thinking, lean execution, and a bias for making things real.
Today, I focus less on training models — and more on architecting full-stack GenAI systems, embedding LLMs into production workflows, and aligning AI capabilities with real-world constraints. That includes everything under the LLMOps umbrella — evaluation pipelines (both human annotations and LLM-as-a-judge), dataset creation for fine-tuning, observability, prompt lifecycle management, and structured experimentation — the kind of stuff that turns a prototype into a production service.
Curious how it all works? Head to "What I did" — real use cases, real numbers, real outcomes.
Let's build something smart, sharp, and scalable together!
Where I've been
AI Data Platform Engineer (Self-employed)
As a self-employed AI Data Platform Engineer at my company, precision&recall, my experience in consultancy greatly helps me connect Product teams with Engineering teams. This skill is crucial for making sure projects run smoothly and successfully. I use a tech stack, including Python, Docker, Terraform, Airflow, PostgreSQL, DuckDB, and AWS services, to build and deploy advanced machine learning models, such as Large Language Models (LLMs), customized to my clients' specific needs. In my role, I also offer strategic advice to make sure machine learning and data science tools fit well into client operations and really help achieve their goals. This involves everything from analysing data and creating visualizations, to train or finetune machine learning models, to setting up and maintaining full MLOps systems that keep models performing well once they're running. My work focuses on turning complex technical details into clear, useful information, helping different teams make better decisions together.
Lead MLOps Engineer
At Vivino, the renowned wine app, I advanced to the role of Lead MLOps Engineer, where I was responsible for optimising the performance and reliability of machine learning deployments. My leadership extended across both technological improvements and team management, focusing on enhancing the scalability and efficiency of our systems through rigorous best practices in coding and infrastructure management. This included the integration of modern tools and methodologies such as FastAPI, Docker, Jenkins, and Terraform, as well as AWS services for robust cloud-based operations. My role was pivotal in transforming our machine learning pipelines and infrastructure, driving faster deployment cycles and higher code quality.
Lead ML Engineer
At Velliv, a Danish pension fund, I served as a Lead ML Engineer, where I developed and deployed various machine learning solutions to enhance data security and process automation. My responsibilities included leading the integration of cutting-edge natural language processing (NLP) technologies and managing a tech stack involving Python, FastAPI, Docker, and various AWS services. I played a key role in mentoring and recruiting staff, overseeing project management, and ensuring the adherence to best coding practices, test-driven development (TDD), and agile methodologies.
External Data Visualisation Collaborator
As a member of the Visualisation Working Group at NASA GeneLab, I assisted in visualising advanced genomics data from all GeneLab studies. This work helped researchers quickly understand the study dataset and the genes significantly impacted by spaceflight. Additionally, it supported NASA GeneLab's goal of empowering citizen scientists to analyze their data. This role was an unpaid, part-time position that I undertook alongside my consultancy job at DAMVAD Analytics. I had the opportunity to visit NASA GeneLab in Boston and also delivered a guest lecture at MIT about some of the visualisation tools we developed.
Senior Data Science Consultant
At DAMVAD Analytics, I started as a Data Science Consultant and was later promoted to Senior Data Science Consultant. In my role, I led a team of eight people and managed important data science projects, helping different clients make better decisions using data. My job involved using tools like Python, Docker, and AWS to build and apply machine learning models. As the technical lead of the Data Science team, I was responsible for recruiting, mentoring team members, managing projects, and speaking at public events. My experience at DAMVAD taught me a lot about managing data and solving problems in various industries.
Data Science Consultant
At DAMVAD Analytics, I started as a Data Science Consultant and was later promoted to Senior Data Science Consultant. In my role, I led a team of eight people and managed important data science projects, helping different clients make better decisions using data. My job involved using tools like Python, Docker, and AWS to build and apply machine learning models. As the technical lead of the Data Science team, I was responsible for recruiting, mentoring team members, managing projects, and speaking at public events. My experience at DAMVAD taught me a lot about managing data and solving problems in various industries.
What people say
As a Product Manager, the main problem I have, is explaining to Engineers the customer perspective, and what value we are trying to achieve with this or that feature so that they can provide the best result. This is never an issue with Sebastian, as he always puts himself in the customer's shoes, testing all the solutions from their perspective and striving to maximize value for the customer. For this reason, I enjoy having brainstorming sessions with Sebastian when it comes to solving new customer problems as he always helps me dig to the root cause.
Another thing I like about Sebastian is his approach to problem-solving, he is not the type of Engineer who goes 'offline' for months to find a solution but is always spiking, testing, iterating, and trying to get some initial results ASAP.
Lastly, I admire how structured he is when breaking down a Jira issue or documenting results, he is always very thorough, also in these matters.
— Stepan Fedorov, Product Manager @ Daltix (Stepan worked with Sebastian on the same team)
Throughout my time with Sebastian, I was consistently impressed with his skills, positive attitude, and relentless focus on creating value for our customers and organization. Sebastian deeply understands machine learning systems, cloud infrastructure, and MLOps, which he leveraged to design, deploy, and manage scalable and reliable ML solutions.
In addition to his technical abilities, Sebastian has a positive and can-do attitude that makes him a joy to work with. He is always eager to help his colleagues and takes ownership of his work, ensuring that everything he touches is of high quality. His excellent communication skills and ability to collaborate with cross-functional teams have made him a valuable asset to our organization.
I highly recommend Sebastian. His technical skills, positive attitude, and commitment to delivering value make him an asset to any organization looking to take its ML solutions to the next level.
— Brian Groth, Chief Technology Officer @ Colourbox & Skyfish (Brian managed Sebastian directly)
I had a chance to work with Sebastian on several projects over the course of the last year. In that relatively short time, his team introduced a lot of best practices into the team's work, completely re-implemented our end-to-end ML pipeline for delivering ML based APIs and under his guidance, his team also developed alternative approaches to optimize our services for performance and cost effectivity. All that work truly brought Vivino's ML engineering to a whole new level.
Sebastian's approach is very methodical, data-driven and with a clear understanding of stakeholder needs. I very much enjoyed our cooperation because we managed to make decisions quickly and move fast with implementation. If given a chance, I'd be happy to work with Sebastian again. Both as a friend and as an engineer.
— Zbyněk Vymazal, Head of DevOps @ Vivino (Zbyněk worked with Sebastian on the same team)
Sebastian is an exceptional leader and a true asset to any team. He possesses a unique combination of creativity and structure, which allows him to effectively lead projects and make decisions. He is a skilled communicator, able to engage with colleagues at all levels and foster a positive and productive work environment. Sebastian is also a thoughtful leader, taking the time to consider the potential consequences of decisions and advocating for his team's ideas and strategies. His work has had a positive impact on the company and his leadership and mentorship have been invaluable assets to the team. I highly recommend Sebastian for any future opportunities and am confident in his ability to excel in any role.
— Claus Vestergaard, Data Scientist @ ZeroNorth (Claus reported directly to Sebastian)
The name is the philosophy
Virtually everyone asks why I named my company precision&recall.
At its core, machine learning relies on precision and recall as two fundamental metrics that measure model performance — precision is about quality (how accurate your positive predictions are), while recall is about coverage (how well you capture all the positive cases).
Generally speaking, think of it like hiring: precision ensures everyone you hire is actually a good fit (quality), while recall ensures you don't miss any great candidates (coverage). These metrics represent the eternal balance in AI: quality versus coverage.
Importantly, my logo visualizes this concept beautifully: two flowing circles that embrace and merge, forming an organic intersection — the gray circle represents precision (quality), the pink circle represents recall (coverage), and their overlap creates the sweet spot where the magic happens. It's that delicate balance between quality and coverage, that fertile ground of possibility, that I constantly optimize for in every AI system I build.
Naturally, the logo also symbolizes collaboration, the convergence of different perspectives, and the idea that the best solutions emerge when seemingly opposing forces find harmony. Even in today's LLM era, this tradeoff remains at the heart of every intelligent system.
Always when I design AI solutions, I'm constantly optimizing for this balance — ensuring systems deliver both quality and coverage, reliability and comprehensiveness. The name reflects my commitment to building AI that achieves both quality and coverage every time.