AnacodicAnacodicAI Labs
← Back to Research

AI-Assisted Rubric-Aligned Assessment for Resource-Constrained Educational Environments

Faculty-advised research — team forming

→ Core contributors included as co-authors

Academic AI

Automated assessment support for resource-constrained educational environments with high enrollment-to-instructor ratios. Feedback consistency and throughput under structural volume constraints are the organizing problem. Research conducted in coordination with active institutional deployments.

RAGNLPLLMPrompt EngineeringSelf-ConsistencyAutomated Grading

The crisis

  • 1 in 8 US teaching positions unfilled or uncertified
  • World needs 44 million new teachers by 2030
  • High-poverty schools: only 39% have AI training vs 67% in wealthy schools (RAND 2025)
  • K-12 teachers delay feedback to weekly/biweekly cycles because workload prevents real-time response
  • Students with disabilities need personalized feedback that overwhelmed teachers physically cannot provide

About this research

The education system faces a structural constraint: teacher workload scales with enrollment, but feedback quality degrades under volume. The core problem is not teacher capability but throughput. Feedback granularity, consistency, and timeliness all suffer at scale. This thread addresses that constraint in resource-constrained institutional environments.

Key findings

  • (In Progress)

Related project

View project →

Apply

Apply by April 30, 2026

Join our Slack community to follow updates and see what we're working on.

Apply Now →

Open roles

ML Engineer

Open

Implement and evaluate RAG evaluation models. Run comparative experiments against manual grading baseline. Extend pipeline to OCR ingestion for handwritten submissions.

Skills: Python, LangChain, Claude API, OCR, Experimental Design

Apply →

Education Researcher

Open

Validate findings against education literature. Map regulatory and accessibility requirements for automated assessment. Support academic writing and citation.

Skills: Education Policy, Research Methods, Academic Writing, Accessibility Standards

Apply →
← Previous: Carbon-Aware Inference Under Deployment Constraints: Extending the CCI FrameworkNext: Multi-Agent Evidence Retrieval with Citation Provenance for Resource-Limited Clinical Settings →