Hi Product Hunt! I’m Moises, a Data & AI Transformation Leader who got tired of shallow stock screeners and decided to build something better. ProspectAI is a multi-agent system that mimics how a professional analyst team works: Research Agent — scans sectors and surfaces candidates using real fundamental data Draft Agent — scores each company across 7 dimensions (momentum, fundamentals, sentiment, risk/reward, and more) Critic Agent — challenges every recommendation, flags R/R violations, and forces revisions Final Agent — produces a clean, structured portfolio with position sizing and action states Built on Claude + CrewAI, with a strict spec-driven architecture based on Openspec framework. Why I built it: Most AI investing tools are wrappers around LLM chat. ProspectAI is an actual pipeline with deterministic rules, schema validation, and a critic loop — the kind of rigor you’d expect from a quant team. Still early, but the pipeline runs end-to-end across Technology, Energy, Semiconductors, and Consumer sectors. Check out the statistics page with live data about the performance of past recommendations. Would love your feedback — especially from anyone who’s tried building agentic financial pipelines. What’s your biggest pain point?

ProspectAI - Multi-agent AI that builds your investment portfolio

Replies