AI Research Agent

Fundamental Research at Machine Scale.

Autonomous agents that read filings, transcripts, and news across 15,000+ stocks — turning unstructured data into revenue, EPS, and return forecasts.

What We Do

Overview

Our AI research agents analyze unstructured financial data with human-like depth — generating and validating investment hypotheses, and surfacing real-time signals on company fundamentals and risks. They cut through the noise to find the variant perceptions others miss, supporting analysts and portfolio managers across the entire research workflow.

How it works

From raw data to a ready-to-act signal.

Every agent runs the same disciplined loop — autonomously, across thousands of companies at once.

01

Data sources

Filings, transcripts, news, and fundamentals across 15,000+ stocks.

02

Hypothesis

The agent forms a testable thesis on where a company is mispriced.

03

Research

It reads the company and its peer group, gathering evidence for and against.

04

Validation

The thesis is tested against fundamentals and quant signals; weak ideas are dropped.

05

Signal / Report

A conviction-scored signal with a written rationale.

Example output

What an agent hands you.

Gradient Deep Research 1:31

Integrations

Fits into how you already work.

Live view

Dashboard

Watchlists, conviction scores, and full rationales in the browser — watch the global agent network cover thousands of equities around the clock.

Explore the live agent network

Further ways to plug in

REST API

Pull signals, forecasts, and rationales straight into your own systems.

Alerts & webhooks

Real-time notifications when a new signal or risk flag appears.

Data export

CSV and feeds that plug into your existing research stack.