AI Pitch Deck Parser with RAG Chatbot & QDrant

This n8n workflow revolutionizes deal flow management for tech startup accelerators overwhelmed by thousands of pitch decks. It uses multimodal LLMs to parse even complex, visually rich PDFs that traditional methods can't handle, transcribing each page into structured markdown. An AI-powered information extractor then generates comprehensive reports on key details like team, market, traction, and funding asks, automatically updating an Airtable database for easy tracking.

Beyond extraction, the workflow uploads transcribed content to a QDrant vector store, enabling a RAG (Retrieval-Augmented Generation) chatbot. This allows users to ask detailed questions about any pitch deck, surfacing precise insights without manual review. It's perfect for scaling evaluation processes, spotting hidden gems among submissions, and collaborating with teams.

Benefits include massive time savings—process hundreds of decks daily instead of manually reviewing them—improved accuracy via AI vision, and enhanced decision-making with on-demand querying. Use cases: accelerators like Y Combinator clones, VCs, angel investors, or corporate innovation teams. Requires duplicating the provided Airtable base and n8n v1.62.1+. Sample Airtable: https://airtable.com/appCkqc2jc3MoVqDO/shrS21vGqlnqzzNUc.