We’re passionate about building AI systems that turn scientific literature into machine-actionable knowledge—structured for AI-first discovery, search, and comparison. Here’s what we’re currently focusing on:
- 🧠 Neuro-symbolic NLP and language models. Developing methods that combine LLMs with structured representations for robust scientific text understanding.
- 🤖 Structured information extraction for scientific knowledge graphs. Extracting and normalizing scientific claims, methods, and results to power AI-assisted search and synthesis.
- 📚 Knowledge organization: schemas, ontologies, and alignment. Supporting semi-automated schema and ontology engineering, including ontology learning and alignment workflows.
- ✅ Evaluation and benchmarking. Building rubrics, benchmarks, and LLM-as-a-judge workflows to assess extraction and synthesis quality reliably.
Interested in collaborating? Reach out at sciknoworg [at] gmail.com.
