SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Albert is building Albert OS, an AI-native operating system for materials science designed to help chemists work more effectively. The company is seeking a Knowledge Architect (Scientific Intent) to join the AI/ML team and define how scientific knowledge is represented, captured, and reasoned over within the platform.
The core challenge: Albert's data model captures structured lab work (materials, quantities, process parameters, transformations, inventory, regulatory context), but the reasoning behind that work lives in narrative and unstructured context—why formulations were revised, what chemists ruled out and why, the thinking patterns that shape processes over months. This role bridges those worlds, making scientific reasoning computable.
Key responsibilities include:
- Define ontologies and taxonomies for scientific intent and process narrative, including types, relationships, and attachment points connecting reasoning to structured experimental data
- Decide how the core data model should be extended, owning reviews that keep new structure traversable, reasoning-friendly, and AI-native
- Figure out how to capture scientific intent as chemists work, including what is worth noting and where annotation can enrich the knowledge representation
- Partner closely with engineers on sophisticated implementations of data models and ontologies, from typed relationships and versioning through production behavior
- Work with process development chemists to understand their reasoning and with AI/ML engineers building on the graph, ensuring structure serves real retrieval and reasoning needs
- Create playbooks and review practices that keep the knowledge layer coherent as more teams write to it
You will work at the schema and data-model level with engineers, co-designing with fluency in graph structure and typed relationships. The role requires comfort with tradeoffs in knowledge representation: what becomes a first-class entity versus an attribute, how to handle schema and versioning over time, cardinality and composition decisions.
Albert has headquarters in the California Bay Area with offices in India, Germany, and Japan; over 50% of employees work outside California. The company operates a hybrid, international culture.
Qualifications and background that fit:
- Motivated by how scientific knowledge gets captured; want defining that to be your primary job
- Can walk through systems you've modeled and speak sharply to tradeoffs (formal knowledge-representation vocabulary not required going in)
- Scientific, industrial, or lab-adjacent background, or strong curiosity about experimental science
- Formal ontology, taxonomy, or knowledge-representation background, or willingness to build one (company will support ramping, including close collaboration with VP, AI/ML)
- Applied ML, numerical modeling, or large-scale data science experience useful, since downstream models and retrieval depend on your structure
- Push back on adding free-text fields when harder structural work is what the problem actually needs
