SI² — SuperIntelligence for Science

AI for Laboratory Operations

Intelligence for the people, records, and decisions that keep scientific laboratories moving.

The mission

Scientific operations.
Intelligent by design.

Build AI tools for laboratory operations that support operational excellence and predictive maintenance. Our research connects diverse operational evidence with grounded reasoning, expert oversight, and useful decision support.

Research stories

The ideas behind the tools.

The ecosystem

A connected research architecture.

01

Operational evidence

Engineering drawings · Incident records · Laboratory documents · Policies & procedures

02

Intelligence foundations

Multimodal retrieval · Retrieval-augmented generation · Agentic reasoning · Evaluation

03

Research & tools

Knowledge discovery · Safety & risk intelligence · Compliance & governance

04

Decision support

Discover evidence · Assess hazards · Review decisions · Share insights

Tools & benchmarks

Explore the research portfolio.

Published
Compliance & Governance

ORCHID

Retrieval-augmented classification of high-risk property with intelligent decision-making.

Research & resources →
Research tool
Compliance & Governance

RegChange

AI compliance tool that compares regulation versions, maps impacts to internal procedures, and generates change recommendations.

Research & resources →
Research tool
Compliance & Governance

PolicyGuard

Extracts requirements from DOE orders & eCFR, compares against internal procedures, and maps gaps and conflicts.

Research & resources →
Research tool
Compliance & Governance

ORImpact

AI decision impact platform that evaluates how operational or policy decisions affect risk posture, compliance, and operations.

Research & resources →
Research tool
Safety & Risk Intelligence

SmartInspect

AI vision tool that analyzes images of any space or equipment with text context to rapidly identify safety hazards.

Research & resources →
Research tool
Safety & Risk Intelligence

CausalAI

Agentic AI performing 5-Whys root cause analysis and barrier extraction on incident reports.

Research & resources →
Research tool
Compliance & Governance

QualityAI

AI anomaly detection for procurement that flags items incorrectly classified as quality-insignificant before they slip through.

Research & resources →
Research outputs

Publications & submissions.

RAG4ReportsPublished

REFSafE: A RAG-Enabled Framework for Predictive Risk Analysis and Automated Safety Report Generation in Mission-Critical Environments

Sanjay Das et al.

Citation

Sanjay Das et al. (2026). REFSafE: A RAG-Enabled Framework for Predictive Risk Analysis and Automated Safety Report Generation in Mission-Critical Environments. RAG4Reports, pp. 47–56.

RAG4ReportsPublished

ORCHID: Orchestrated Retrieval-Augmented Classification of High-Risk Property with Intelligent Decision-Making

Sanjay Das et al.

Citation

Sanjay Das et al. (2026). ORCHID: Orchestrated Retrieval-Augmented Classification of High-Risk Property with Intelligent Decision-Making. RAG4Reports, pp. 57–64.

WACV 2027Accepted

BLUEPRINT Rebuilding a Legacy: Multimodal Retrieval for Complex Engineering Drawings and Documents

Ethan Seefried et al.

Citation

Ethan Seefried et al. (2027). BLUEPRINT Rebuilding a Legacy: Multimodal Retrieval for Complex Engineering Drawings and Documents. WACV 2027.

JCDLPublished

ChemSafe: Hybrid Retrieval-Augmented Decision Support for Laboratory Chemical Hazard Assessment

Ran Elgedawy et al.

Citation

Ran Elgedawy et al. (2026). ChemSafe: Hybrid Retrieval-Augmented Decision Support for Laboratory Chemical Hazard Assessment. JCDL.

ChemRxivPreprint

SafeChem: A Benchmark Dataset for Multi-Label Chemical Hazard Prediction and LLM Safety Hallucination Evaluation

Ran Elgedawy et al.

Submitted to NeurIPS

Citation

Ran Elgedawy et al. (2026). SafeChem: A Benchmark Dataset for Multi-Label Chemical Hazard Prediction and LLM Safety Hallucination Evaluation. ChemRxiv.

JCDLPublished

SafeLens: AI-Augmented Exploratory Analysis over Organizational Safety Records for Actionable Insights Generation

Sanjay Das, Ryan Burchfield, Gavin Wiggins, Alan Guy, and Tirthankar Ghosal

Citation

Sanjay Das, Ryan Burchfield, Gavin Wiggins, Alan Guy, and Tirthankar Ghosal (2026). SafeLens: AI-Augmented Exploratory Analysis over Organizational Safety Records for Actionable Insights Generation. JCDL.

SIGDIALPublished

Enhancing Operational Safety via Agentic Dialogue Hazard Identification Analysis

Sanjay Das, Ran Elgedawy, Ethan Seefried, Ryan Burchfield, and Tirthankar Ghosal

Citation

Sanjay Das, Ran Elgedawy, Ethan Seefried, Ryan Burchfield, and Tirthankar Ghosal (2026). Enhancing Operational Safety via Agentic Dialogue Hazard Identification Analysis. SIGDIAL, pp. 757–772.

Browse AI for Laboratory Operations publications →