Cortex-BRD
Link to open source: https://cortex-brd.vercel.app/
We propose CORTEX-BRD, an AI-powered Business
Requirements Intelligence System that automatically transforms
scattered business communications into structured, audit-ready
BRDs. The system ingests multi-channel data including emails
from the Enron Email Dataset and meeting transcripts from the
AMI Meeting Corpus. Using advanced NLP and large language
models, it filters noise, detects project-relevant discussions,
extracts requirements, decisions, stakeholder inputs, and
timelines, and maps them into a temporal knowledge graph.
Unlike traditional summarizers, CORTEX-BRD provides full
traceability every requirement links back to its original email
thread or meeting timestamp. It also analyzes stakeholder
influence and tracks requirement evolution over time. The system
generates structured, explainable BRDs while supporting human
validation, ensuring accuracy, accountability, and enterprise
grade reliability.
This build was uploaded as a hackathon project
.jpg)