Loading...
Discovering amazing AI tools

A benchmark dataset and evaluation suite mapping Product Hunt launches to Series A outcomes for predictive modeling of startup funding.
A benchmark dataset and evaluation suite mapping Product Hunt launches to Series A outcomes for predictive modeling of startup funding.
PHBench is a research benchmark and dataset that maps Product Hunt launches to verified Series A funding outcomes within 18 months, enabling predictive modeling of startup fundraising. The dataset covers 67,292 featured Product Hunt posts (2019–2025) linked to 528 confirmed Series A outcomes and provides extensive engineered signals per post. PHBench includes 61 engineered features (engagement, rank, maker, temporal, topic flags, interaction terms), standard train/validation/test splits with withheld test labels for blind evaluation, and accompanying ML and LLM experiment baselines described in an arXiv paper. The benchmark is intended for researchers and practitioners building classifiers, ranking systems, and graph-based models to prioritize or study early-stage investment signals; access is governed by dataset license/conditions and submission procedures for scoring are handled via Vela Partners.
