Machine Learning Research Manager, Miro

• Proven track record of technical leadership: 2+ years of experience managing high-performing Applied Science or ML Engineering teams in a product-led tech company or top-tier research lab
• Multimodal & GenAI Depth: You don't just use APIs; you understand the architecture of Transformers, Diffusion models, and Graph Neural Networks (GNNs). You can guide a team through the complexities of fine-tuning and RAG at scale
• A "Product-First" Researcher: You understand that accuracy metrics (F1, AUC) are proxies, not goals. You prioritize user value and latency constraints in production
• Curiosity for the "Unsolved": You are excited by the ambiguity of modeling "collaboration." How do you quantify a "good brainstorm"? How do you autocomplete a flowchart?
• Strategic Communication: You can articulate the difference between "hype" and "utility" to executive stakeholders and align research efforts with Miro’s long-term strategy
• Option A: PhD in Computer Science, Statistics, Mathematics, or related field + 2+ years of people management experience
• Option B: Master’s degree or equivalent deep technical experience + 5+ years of industry experience in ML, including 2+ years of people management

Desirable

• Miro is the online workspace for innovation, used by 100M+ people to build the next big thing. We are seeking a Research Manager to lead the team defining the brain behind the "Intelligent Canvas."
• This is not for building a standard GenAI or Recommendation engine. You will lead a team operating at the intersection of Vision, Language, and Graph Theory, working with a dataset unlike any other in the industry: spatial, unstructured, and deeply human collaboration data
• You will bridge the gap between open-ended research (LLMs, Diffusion Models, GNNs) and product impact, empowering teams to dream, design, and build faster
• Build and lead a world-class applied research team, hiring and mentoring researchers who excel at the intersection of deep learning theory and production engineering
• Define the research roadmap to support the "Intelligent Canvas," identifying opportunities to model complex user behaviors—from multi-user collaboration on an infinite canvas, to multi-format AI-powered generation (e.g. slide deck, technical diagram, web app prototypes, etc.), and more
• Pioneer research on unique spatial datasets. Unlike standard text/image corpuses, you will explore massive multimodal and graph-based datasets, uncovering how teams organize information spatially and collaborate together to solve complex multi-modal problems
• Drive the "Research-to-Product" velocity. You will create the framework for rapidly testing foundation models (e.g., GPT-4, Llama, Stable Diffusion) and fine-tuning them for specific domain tasks (e.g. prototype generation, diagram generation, mindmap generation)
• Cultivate a culture of scientific rigor. Encourage the team to stay at the cutting edge (NeurIPS, CVPR) while maintaining a relentless focus on shipping features that delight users
• Partner with Engineering & Product Leadership to translate abstract AI capabilities into intuitive solutions that feel like magic to our users
• Architect organizational processes for model governance, ensuring rigorous evaluation frameworks, reproducibility, and ethical AI practices