📄 RAW INPUTS

📄 MCA Draft Document

📄 Stakeholder Comments

Pre-Processing (clean text: tokenize, lower-case, remove HTML)

NER / KeyBERT (keyphrase extract)

Word-Cloud Gen (wordcloud lib)

📄 Word Cloud Data

📄 Keyword Frequencies

📄 Clean Text

Hugging Face stance detection

📄 BART/TS etc (uses text + sentiment)

📄 Summaries

📄 Sentiment Results

Database / API Layer

📄 Downloadable Reports

architectural diagram