Get Started

Welcome

SonicAlpha turns public-market podcasts into structured, searchable investment insights.

We identify relevant conversations, extract individual investment claims, and attach the context needed to use them: who said it, which companies it relates to, sentiment, publication time, and the underlying source.

How it works

Most integrations start with two endpoints:

Ask

POST /v1/insights/ask

Ask a research question in natural language.

SonicAlpha searches across the corpus and returns the most relevant evidence. It can also synthesize that evidence into a cited answer.

POST /v1/insights/search

Search the underlying insight corpus directly using queries and filters such as ticker, speaker, sentiment, podcast, or date.

Use Search when you want to handle ranking, orchestration, or synthesis yourself.

The data

SonicAlpha organizes the corpus into three main objects:

Podcast → Episode → Insight

An insight is the core object: a discrete investment-relevant claim tied back to the speaker and source where it came from.

Start with the Quickstart to make your first request.