AI Persona Generation from Survey Data | AI Hackathon

Watch a complete walkthrough of an AI pipeline that converts 500+ lunch-combo survey responses into four evidence-based consumer personas. Recorded during an AI Hackathon, this demo pairs psychographic scoring with LLM narrative framing for actionable market insights.

This session documents a reproducible workflow that transforms a skewed, Ontario-heavy dataset into four synthetic personas ready for product, design, and marketing teams. The process integrates HEXACO and Big Five trait extraction, demographic cross-tabs, and large-language-model storytelling to connect raw data with strategy.

What You’ll Learn
Data Audit – Identifying and qualifying sample bias before analysis.

Psychographic Mapping – Deriving openness, conscientiousness, agreeableness, and related traits from attitude items.

Persona Construction – Structural Absolutist, Broad Inclusionist, Contextual Pragmatist, and Intentionalist.

Applications – Menu design, feature prioritization, brand narrative, and retention modeling.

The GitHub repository contains the full codebase, prompt templates, and documentation. Fork the project, replace the CSV, and generate personas for your own dataset.

Repository, dataset, and slide links below.

#AIPersonaGeneration #SurveyData #HEXACO #BigFive #AIHackathon #MarketResearch

📍 Chapters
0:00 – Introduction and dataset overview
0:28 – Sample bias and limitations
0:57 – Extracting psychographic features (HEXACO & Big Five)
1:27 – Persona 1: Structural Absolutist
1:50 – Persona 2: Broad Inclusionist
2:12 – Persona 3: Contextual Pragmatist
2:32 – Persona 4: Intentionalist
2:55 – Synthetic personas vs. raw demographics
3:15 – Product and marketing use cases
3:45 – Roadmap and community contributions