Applied Research Scientist - Digital Enterprise Transformation

Do you want to push the boundaries of machine learning and scale it to users across the world? Do you wish for a product centric work environment, where machine learning and analytics drive action – not just insight?

With a strong focus on the applied machine learning research discipline, you will take part in innovation and design of new products, as well as tweak and mature existing ones. You will join a team of highly talented data scientists and collaborate with several product teams consisting of product owners, backend- and frontend developers and DevSecOps specialists. You will be pushing the applied research agenda, aiming to create and continuously innovate the ML systems of tomorrow.

Join a department focused on our most valuable digital products

You will join the Research and AI Lab team in Digital Enterprise Transformation. Our new Research and AI Lab unit is characterized by an intrapreneurial mindset, and is collaborating holistically with established MLOps, DevSecOps and Data Science teams that have had ML systems in production on a global scale for several years. We strive to base our work on knowledge and insight rather than hierarchical structures, and we make sure that our decisions are based on conversations between people with different competencies rather than one individual.

Examples of what we have worked on in 2024:

• Trained, tuned, and served large generative language models for document processing and for conversation using both text and multimodal data.
• Designed a new large machine learning system set for release in 2025.
• Implemented and deployed a custom ‘attention-free’ transformer model, including functionality for large-scale model training and inference.

Current and ongoing plans set for 2025:

• Establish industrial research focus by driving research-based project work and collaboration with academic research institutions, including publishing our work in relevant ML journals/conferences
• Planning and forming our journey into “Agentic AI” based methods when applicable
• Generally: Become creators rather than consumers in GenAI and multimodal machine learning.
• And more!

You thrive when solving highly complex challenges

The use cases we handle tend to have a high degree of complexity, requiring non-deterministic problem solving (i.e., the use of ML/AI), near real-time data processing, a need for high availability, vertical and horizontal scalability, and a very high volume of transactions. Solving complex problems with the use of machine learning, you need an open and analytical mind, finding pragmatic solutions that balance between the theoretical standpoint and what is possible.

Be the driver of our AI/ML and GenAI journey

As our Applied Research Scientist, your work will be multifaceted. We will work on our research agenda and push for new machine learning applications and methods, while ensuring this work is well-adopted by the existing teams for use in downstream products that will directly impact DSV’s divisions. You will have a mix of independent work and participate in existing project-focused teams. You will take part in promoting our AI efforts internally and have the opportunity to promote our work externally, e.g. as a speaker at partner events.

Responsibilities:

• Advance GenAI and multimodal research and development: Spearhead applied research in GenAI, ‘Agentic AI’ and multimodal machine learning, encompassing image, layout, and text/language modeling.
• Drive applied research that can be applied to DSV’s product development. Drive cross-project innovation to tackle complex challenges and real-world problems to gain new insights and apply findings to enhance model performance and outcomes.
• Foundational model development: Independently architect, train, and evaluate foundational models to advance GenAI LLMs and multimodal systems in DSV’s field of work.
• Help keep the wider data science group in DSV and keep yourself up to date with the latest developments in the field.

Skills and experience we value and would love you to have already:

• Educational Background: Preferably a PhD or master's degree in CS, CE, ML or a related field, with proven research experience.
• Strong Theoretical Foundation: A comprehensive skill set in mathematics, statistics, and algorithms is essential for implementing research ideas, building upon existing research to create new, and adapting established work to our domain.
• Software Development: A good understanding of core programming and development concepts: data structures, algorithms, and object-oriented programming. You will both be performing rapid prototyping and contribute to large collaborative machine learning codebases. We primarily use Python due to the high availability of machine learning frameworks.
• Publications and Open-Source Contributions: Proven track record of first-author peer-reviewed publications that demonstrate your ability to lead research projects and navigate the publication process, along with contributions to open-source projects.
• Model Development and Architecture: Proficiency in building from scratch, modifying and optimizing existing models and in advanced model architectures, particularly for GenAI and multimodal applications.
• Accelerated and Distributed Compute: Proficiency in accelerated parallel computing, including scalable data processing and leveraging GPU technology for enhanced performance and model-training scalability. We use frameworks such as: PyTorch, Kedro, Apache Beam, and more.
• Continuous Learning: Commitment to staying current with the latest advancements in machine learning and applying new academic insights to solve complex business challenges.
• Practical Experience: Hands-on experience with large-scale, real-world machine learning projects.
• Self-Starter & Ownership: We're looking for someone who is proactive, takes initiative, and can drive projects independently. You should be comfortable taking ownership of objectives and pushing things forward.

We understand that no candidate may meet every single requirement listed here. If you meet some of the qualifications, but not all, we still encourage you to apply! We're looking for passionate individuals who are eager to grow and contribute to our team.

What we offer to you:

• Direct Communication: We engage directly with VP and C-level stakeholders, ensuring your insights and contributions are heard at the highest levels.
• Influence on MLOps: Contribute indirectly to the development of our robust, in-house MLOps platform, collaborating with a dynamic and expanding MLOps team dedicated to supporting our mission.
• Global Scale Experience: Leverage our extensive experience with ML systems in production on a global scale, benefiting from our experience in development and operations processes for large-scale machine learning systems.
• A proven team: A group which is already self-financed based on realized returns of machine learning business cases.
• Research and AI Lab: Be part of our Research and AI Lab initiative, where you can influence our: Methodologies, direction, and research agenda.

If you want to apply We will be happy to answer any questions you may have regarding the position and about your options in DSV. You are welcome to call Ramus Kær Jørgensen on +45 25452587.

We look forward to receiving your application as soon as possible. We will process the applications as we receive them.

DSVGlobalIT