Company Description
Articul is an Amsterdam-based scientific intelligence lab specializing in advanced decision systems for enterprise leaders. By employing interdisciplinary scientific methods, we deliver tailored, research-to-deployment projects designed to unlock competitive advantages for clients in volatile business environments.
Our team of scientists, engineers, and analysts work collaboratively across domains to extract value from underutilized data and provide actionable, defensible insights. Through our service portfolio (Atlas, Synthesis, and Origin) we help organizations minimize uncertainty and maximize returns on critical decisions.
Role Description
Learn to research, test, and experiment with open-weight and frontier language models to solve real methodological challenges in quantitative and qualitative research. You'll work at the intersection of AI capabilities and research rigor, testing whether models can reliably extract signal from unstructured data, augment causal inference workflows, or accelerate literature synthesis without sacrificing validity.
Your work will involve systematic experimentation: designing protocols to evaluate model performance on research tasks, building prompt architectures that maintain methodological standards, and creating reproducible pipelines that move from proof-of-concept to operational tools. You'll contribute to internal knowledge systems that document what works, what fails, and why, building institutional memory around AI application that most organizations lack.
Role Responsibilities
- Systematic evaluation frameworks for LLM performance on research tasks (coding qualitative data, extracting structured information, synthesizing evidence)
- Production-ready tools that embed AI into Articul's research workflows while maintaining scientific standards
- Internal knowledge systems documenting best practices, failure modes, and decision frameworks for AI application
- Experimental protocols testing novel applications of multimodal models, retrieval systems, and agentic architectures
What We Offer
This is an unpaid position. What we offer instead is work on real client projects with actual stakes, instead of simulated exercises. You will build tools and systems that get used beyond your internship and leave with a portfolio showing what you have created, not just what you have observed. We give you creative freedom and the expert support needed to develop your ideas.
We believe in compound learning. You will develop practical expertise in AI application, research methods, and systematic problem-solving. There is no predefined playbook in this domain, so building judgment and learning to operate in ambiguous environments is central. This is an increasingly sought after skill set.
Additionally:
- Small, agile team of high-performing scientists, engineers, and analysts
- Horizontal organizational structure with substantial decision-making authority
- Dedicated weekly hours for individual learning and research
- Collaborative environment across disciplines with access to diverse expertise
- Flexible working hours with remote-first policy
Qualifications
Personal Attributes:
- High curiosity and intrinsic motivation
- Demonstrated agency and comfort with ambiguity
- Creativity in problem-solving
- Grit and conscientiousness in execution
- Preference for transparent work
- Interest in research, engineering, business, technology, or AI
Preferred but not Required Technical background:
- Experience with Python, APIs, or ML frameworks
- Prior work with LLMs (prompt engineering, context-engineering, RAG)
- Research methods knowledge (quantitative or qualitative)
- Technical coursework (programming, data science, statistics) or equivalent self-taught skills
Minimum Requirement: 2nd year Bachelor's or beyond, any discipline, C1 level in English or above
Location: Remote with possibility for in-person collaboration in Amsterdam
Duration: Between 3 and 6 months, 8 hours per week, beginning in March