The Product

Conversational Research, Fully Cited

AI Chat brings Kaleidoscope's vector database into a natural-language interface purpose-built for financial research.

Your Research Stack in One Conversation

Instead of building complex queries or scanning raw filings, you just ask. "Which biotech companies filed 8-Ks about clinical trial results this quarter?" AI Chat retrieves the answer from our 50M+ document index, structures it for you, and cites every source so you can verify and drill deeper.

Built entirely in-house, AI Chat layers an LLM purpose-tuned for financial research on top of the same vector database that powers our semantic search. The result is a chat experience that actually knows what a Form 4, an 8-K Item 1.03, and a 13F manager classification are.

Natural-Language Questions – No query syntax, no boolean operators. Ask the way you think.
Cited Sources – Every answer links to the specific filing, section, and sentence that grounded it.
Multi-Turn Conversations – Follow up, narrow down, or pivot — the context carries through.
50M+ Document Index – SEC filings, news and press releases, and 13F institutional data in one place.

AI Chat interface showing a research conversation with cited sources

Features

What Makes It Different

Natural-Language Input

Ask about companies, events, risks, or themes the way you would ask a colleague. No special syntax required.

Verified Citations

Every claim is grounded in a source — filing accession number, section, and exact sentence. Click through to verify instantly.

Cross-Data Coverage

SEC filings (10-K, 10-Q, 8-K, DEF 14A), news and press releases, and 13F institutional positions — all searchable in a single query.

Multi-Turn Follow-Ups

Narrow a result set, compare two companies, or pivot to a different angle — the conversation context carries forward automatically.

Use Cases

Who Uses AI Chat

Research Analysts

Screen for companies mentioning specific themes in filings, pull comparable disclosures across a peer group, and monitor portfolio companies for material language changes — all in a single conversation.

Investor Relations Teams

Draft Q&A scripts grounded in your own filing language, benchmark your disclosures against peers, and track how your risk factor language has evolved quarter over quarter.

Legal & Compliance

Search across thousands of filings for precedent disclosure language, identify how companies have handled similar risk factors, and flag unusual deviations in your clients' filings.

Demo

See It in Action

A sample conversation — the kind of research that used to take hours, now in seconds.

Kaleidoscope AI Chat

U

Which semiconductor companies had significant insider buying in the last 30 days alongside rising positive echo signals?

K

I screened SIC 3674 (semiconductors) for insider buying and positive echo signals in the last 30 days. Three companies stand out:

ONTO (Onto Innovation) — CEO purchased 10,000 shares at $172. Positive echo 0.74, profile mutation 0.31. Latest 10-Q flagged expansion of backlog language.
CRUS (Cirrus Logic) — CFO and two directors bought within 3 days of each other. Positive echo 0.69.
AOSL (Alpha and Omega) — Director purchase. Echo composite elevated; positive echo 0.61.

Sources: Form 4 filings (EDGAR), ONTO 10-Q 2026-03-15, CRUS 10-Q 2026-02-28

U

What's the specific language in ONTO's 10-Q about backlog?

K

From ONTO's 10-Q filed 2026-03-15, MD&A section:

"Backlog at the end of the quarter was $218.4 million, an increase of 34% compared to the prior quarter, reflecting strong customer pull from leading-edge advanced packaging programs and continued demand for our metrology solutions in gate-all-around process nodes."

ONTO 10-Q · Accession 0001234567-26-000123 · MD&A, p. 24

AI Chat for SEC Research