APAKSHAT PORWAL
All projects2026 · Active development

Gradia

An AI learning platform that turns uploaded source documents into structured lesson assets.

Repository
PRODUCT // CAPTURE
Gradia world map for navigating learning regions.
Gradia active-recall study hub with grouped reviews.
Gradia learner profile with level progress and skill chart.
PRODUCT // VIEWS

Inside Gradia

Gradia world map for navigating learning regions.
Gradia world map for navigating learning regions.
Gradia active-recall study hub with grouped reviews.
Gradia active-recall study hub with grouped reviews.
Gradia learner profile with level progress and skill chart.
Gradia learner profile with level progress and skill chart.
Gradia daily-focus dashboard and current lesson route.
Gradia daily-focus dashboard and current lesson route.
Gradia learning-domain selection screen.
Gradia learning-domain selection screen.
Gradia regression lesson route and module summary.
Gradia regression lesson route and module summary.

The problem

Turning large collections of PDFs and DOCX files into consistent learning material is slow, difficult to validate, and prone to rendering failures.

The approach

Orchestrate document extraction, semantic tagging, generation, validation, and persistence as an asynchronous multi-stage content pipeline.

My contribution

  • Built the end-to-end pipeline for transforming uploaded documents into structured lesson assets.
  • Designed PostgreSQL/RDS schemas for metadata, semantic tags, version history, and generated outputs.
  • Added automated validation across multi-stage LLM workflows to improve reliability.

What exists now

  • Processed more than 1,000 documents and asynchronous workloads exceeding 1,000 pages.
  • Reduced manual lesson creation time by 80%, rendering failures by 40%, and average processing time by 35%.
Next projectLearn Lens