Building open, auditable infrastructure for computational toxicology and next-generation risk assessment (NGRA).
I work at the intersection of:
- computational toxicology
- agentic scientific workflows
- tool-augmented LLM systems
- interoperable scientific infrastructure
- evidence, PBPK, AOP, QSAR, and exposure tooling
- ToxMCP Suite — umbrella repo for the ToxMCP ecosystem
- PBPK MCP — structured PBPK execution and reporting workflows
- AOP MCP — mechanistic pathway discovery, mapping, and draft authoring
- CompTox MCP — evidence federation for EPA CompTox
- O-QT / QSAR Toolbox tooling — AI-assisted interaction with OECD QSAR Toolbox workflows
Most toxicology workflows are fragmented across databases, tools, models, and reporting surfaces.
I build modular MCP-based infrastructure that makes these workflows more:
- auditable
- inspectable
- interoperable
- automation-ready
- ToxMCP preprint — bioRxiv DOI:
10.64898/2026.02.06.703989 - O-QT paper — Computational Toxicology (2025)
- Website — in4r.ai
- ORCID — 0000-0002-7637-8515
- evidence-aware MCP servers for toxicology
- PBPK and exposure workflow infrastructure
- reproducible, machine-readable reporting
- open scientific tooling for NGRA
Interested in collaborations around:
- toxicology infrastructure
- scientific agent systems
- interoperable research tooling
- open-source applied AI for science
