Research Product Manager — Structured AI Systems

📋 Description

• • Granica Inc. is at the forefront of AI research and infrastructure, specializing in reliable and steerable representations for enterprise data. Our flagship product, Crunch, acts as a policy-driven health layer, ensuring large tabular datasets are efficient, reliable, and reversible. Building upon this robust foundation, we are developing Large Tabular Models – sophisticated systems designed to learn intricate cross-column and relational structures, thereby delivering trustworthy answers and automation with inherent provenance and governance.
• • We are seeking a highly motivated and technically adept Research Product Manager to join our Research & Applied Systems team, based in our Downtown Mountain View, CA office. This is a critical, office-based role requiring five days a week of in-person collaboration, essential for the deep, cross-disciplinary work involved.
• • The core mission of this role is to drive the coherent, efficient, and scalable advancement of Granica's foundational research efforts. These efforts span critical areas such as tabular data learning, large tabular models, structured and relational representation learning, compression-aware and efficiency-driven AI, hybrid symbolic, relational, and neural systems, and the intricate intersection of information theory, learning theory, and large-scale systems. Crucially, these research initiatives are directly integrated with real-world production systems handling petabytes of enterprise data.
• • The Research Product Manager will act as a vital connective tissue, linking people, ideas, compute resources, and systems to transform breakthrough research into durable, impactful capabilities. This is not a traditional program management role; it demands a unique blend of technical understanding and strategic vision.
• • The ideal candidate will possess a deep understanding of how large AI models are trained, deployed, and maintained within production environments. They will be adept at translating foundational modeling advancements into tangible economic value for enterprise infrastructure. Furthermore, they will play a pivotal role in shaping both the technical execution roadmap and the overarching economic strategy that underpins our research endeavors.
• • **Key Responsibilities and Ownership Areas:**
• • **Productionization of Structured AI Models:** Collaborate closely with Research and Systems teams to architect the lifecycle of large tabular models. This includes designing training methodologies for data formats like Parquet, Iceberg, and Delta; defining the requirements for training infrastructure, encompassing data pipelines, distributed training frameworks, and robust evaluation loops; specifying inference architectures (batch vs. streaming, embedding materialization, retrieval strategies); establishing maintenance protocols such as retraining cadences, data drift detection mechanisms, and schema evolution management; and critically analyzing storage and compute trade-offs within real-world systems. You will be expected to possess a strong grasp of data layout, compute scheduling, model lifecycle management, infrastructure bottlenecks, and evaluation pipelines.
• • **Economic Value Extraction:** Contribute significantly to defining the economic landscape of our AI offerings. This involves identifying target buyers (e.g., infrastructure teams, ML teams, data platform teams), pinpointing areas where economic value can be unlocked (e.g., through compression, compute savings, enhanced model accuracy, improved governance), developing methodologies for quantifying this value (e.g., cost curves, workload modeling, infrastructure substitution analysis), and strategizing the conversion of research capabilities into revenue streams and sustainable platform advantages. A strong intuition for enterprise infrastructure economics is paramount.
• • **Bridging Research to Durable Systems:** Actively identify and champion modeling advances with significant production potential and economic viability. Conversely, you will be empowered to discontinue research directions that lack practical system or economic feasibility. You will define clear integration paths for research outputs into existing enterprise workloads and collaborate directly with the Chief Research Scientist on prioritizing the research agenda.
• • **This role is explicitly NOT:** A coordination-heavy research program manager, a product manager focused on consumer AI personalization, or a purely academic researcher. It requires a hands-on, strategic approach to productizing cutting-edge AI research.
• • Granica offers a unique opportunity to work at the intersection of fundamental research and enterprise impact, shaping the infrastructure that defines how efficiently the world can create and apply intelligence. You will have real ownership and influence, deeply connected to the company's core mission. We are building a generational company with an enduring horizon, backed by leading investors and luminaries, focused on decades of impact rather than short-term product cycles. Join us to build the foundational data systems that power the future of enterprise AI.