Product · API

Cited research,
inside your product.

The same engine behind the CalQuity platform, exposed as an API. Drop a conversational research assistant into your own site or app — one that reads the sources, reasons in steps, and cites every answer, so your users get research-grade responses without ever leaving your product.

Book a demo See how it works

acme-research.com · Assistant

Cited

Which segment drove the revenue beat, and did margins hold?

The beat was led by Cloud — cloud revenue grew the fastest¹ while margins expanded².

Sources · every claim traced

1
Q3 FY26 Results · p.12
…Digital services revenue rose, outpacing all other lines…
2
Earnings call · CFO
…operating margin expanded versus the prior quarter…

What you can build

A research assistant, on your terms.

Everything that makes CalQuity defensible inside the platform — full-corpus retrieval, step-by-step reasoning, and a citation on every claim — available as an API you compose into your own experience.

Conversational research assistant

A multi-turn assistant your users talk to in plain language. It decomposes the question, searches across your sources, and reasons in steps — research-grade answers without leaving your product.

A citation on every claim

Each answer traces back to the exact source paragraph it came from. Your users verify in one click, and you ship an assistant that never asks to be taken on faith.

Grounded in your domain

Point it at your corpus — filings, transcripts, reports, your own documents. Answers stay anchored to what you connect, tuned to the vocabulary your users actually use.

Streaming, structured responses

Tokens stream as they are generated, and citations arrive as structured JSON alongside the text — so you can render footnotes, source cards, and links however your UI calls for.

Integrate

A few lines to a cited answer.

Install the SDK, point it at your corpus, and you’re a few lines from a cited answer. TypeScript and Python, fully typed — or call the REST endpoint directly. Either way the answer streams back with every citation as structured data you can render however your interface needs.

Drop-in SDKs for TypeScript and Python — or plain REST
One call — query in, answer and citations out
Citations as typed, structured data, not buried in prose
Streaming by default

your-app.ts · @calquity/sdk

SDK

// npm i @calquity/sdk
import { CalQuity } from '@calquity/sdk'

const cq = new CalQuity({ corpus: 'filings' })

// one call — answer + citations, streamed
const { answer, citations } = await cq.ask(
  'Which segment drove the beat?'
)

Returns

answer: “Cloud led the beat…”·2citations

Built for domain-specific research

For platforms where the answer has to be right.

The API is built for surfaces where a confident-but-wrong answer is a liability — research portals, data platforms, and knowledge products where every claim has to trace to a source.

Equity research portals

Let subscribers ask questions against filings, transcripts, and your analysts’ notes — and get a cited answer instead of ten links.

Financial data platforms

Turn the data you already license into a conversational layer that explains, compares, and cites the figures behind every response.

Knowledge & document products

Add a research assistant over a large, domain-specific corpus — contracts, policies, manuals — with answers traceable to the page.

Customer-facing experiences

Give your users a research assistant inside the product they already use, lifting support and self-serve research without a new tab.

Book a demo

Put cited research in your product.

Bring your corpus and a question your users would ask. We’ll wire up the API and show a live cited answer inside a sample of your interface.

Book a demo How the engine works