The OpportunityIn SDC (Search, Discovery, and Content AI), we power semantic search, recommendations, and agentic experiences across Adobe's creative ecosystem.
A core priority is
knowledge-grounded AI - integrating knowledge graphs, multimodal embeddings, LLMs, and ranking systems to deeply understand intent and creative content.
We're hiring a
Principal Scientist to define how structured knowledge and foundation models work together at scale. You will architect next-generation graph-driven search, hybrid retrieval, and grounded generative systems that power real products used by millions. This is a deeply technical, high-ownership role!
What You'll DoYou will build and evolve large-scale knowledge-grounded systems that connect queries, entities, concepts, and multimodal signals into coherent semantic architectures.
Your work will span:- Hybrid neural-symbolic retrieval and ranking
- Structured intent modeling and entity grounding
- Multimodal representations aligned with knowledge graphs
- Graph-grounded RAG and agentic systems
You will also raise the scientific and architectural bar by:
- Defining evaluation frameworks for intent accuracy, entity grounding fidelity, graph coverage, and semantic relevance
- Establishing modeling and experimentation standards across teams
- Driving principled system building rooted in measurable impact
- Mentoring senior scientists and crafting cross-org technical direction
Expect ambiguous problems, deep technical challenges, and visible product impact!
Scope & ImpactAt the P60 level, you will build semantic architecture across multiple products and set the strategic direction for knowledge grounded AI. You will architect how our knowledge-graphs and foundation models evolve in tandem. You will be a recognized authority in knowledge-centric AI systems.
Basic Qualifications- PhD or equivalent experience (preferred) or MS in Computer Science, AI, ML, or related field
- 10+ years building and deploying large-scale AI/ML systems
- Deep expertise in knowledge graphs, information retrieval, NLP/intent modeling, multimodal learning, or LLM/RAG systems
- Proven track record delivering production-grade semantic platforms
Preferred Qualifications- Experience designing ontologies or evolving domain taxonomies
- Familiarity with graph embeddings, GNNs, or hybrid graph-LLM architectures
- Experience with entity linking, semantic parsing, or large-scale indexing systems
- History of cross-organizational technical influence
What Sets You ApartYou think in systems, not isolated models.
You hold a high bar for rigor, clarity, and measurable impact.
You influence through technical depth and sound judgment.
You care about correctness - and building things that last!
