AI chat

Ask about any stock.
Get the filing, not a guess.

It reads 10-Ks, earnings calls and insider trades for you — and shows exactly where every answer came from.

7-day free trial · Claude Sonnet 5, GPT-5.6 and Grok 4.6 built in

Insighthread News with a Rubrik story open: CEO Bipul Sinha plans to sell 500,000 shares under a pre-paid forward contract with Goldman Sachs, tagged Insider SellChat composer with @RBRK News attached, asking: “Is the CEO dumping the stock, or is this a planned sale?”

No copy-pasting

It already knowswhat you’re looking at.

Open chat from any company, story, filing or earnings call. It’s already attached — just ask.

News page with a Rubrik insider-sale story open, and the AI chat asking about @RBRK News: “Is the CEO dumping the stock, or is this a planned sale?”

What it can do

Pick a question.
Watch it work.

It opens the right filing, runs the numbers and answers in plain English — with the source attached.

  • Reads any filing
  • Watches the insiders“Check the QXO insider trades”
  • Pulls the financials
  • Runs your watchlists
  • Scans news across a list

Check the QXO insider trades

Get insider activity summary · 13s

  • Get insider activity summaryQXO

QXO Insider Trading — Last 180 Days

No open-market buys or sells (Form 4 codes P/S) were reported for QXO in the past year. All 24 recent Form 4 filings are equity awards and exempt transactions — grants and vesting under compensation plans, not market purchases or sales.

Key takeaways

  • • No cash transactions: every entry is a non-cash equity grant, not an open-market trade.
  • • No sales, no purchases: zero P or S codes in the trailing year — no signal of insiders putting money in or taking it out.
  • QXO Form 4

Shows its work

Every answer
shows where it came from.

Identify all meaningful read-across for @NVIDIA from Alphabet, Microsoft, and Amazon’s recent earnings reports and transcripts

Across the three hyperscalers, the read-across for NVIDIA is consistently positive on demand and mixed on timing.

Alphabet raised FY capex guidance to $75B, explicitly citing AI infrastructure. Microsoft flagged that it remains capacity constrained through at least H1, and Amazon pointed to accelerating AWS growth driven by AI workloads.

The common thread is that demand is not the constraint — supply and power availability are.

  • GOOGL Q3 10-Q
  • MSFT Q1 call
  • AMZN Q3 10-Q
@AAPL Transcript

What does that imply for NVDA’s Q4 guide?

If supply is the binding constraint rather than demand, the guide is more a function of allocation than of end-market strength. Watch the data-center segment commentary and any change in supply-chain language.

Here are the two decks — pull the capex figures out of both

Both decks put FY capex above prior guidance. Alphabet moves to $75B from $52B, and Microsoft's slide implies roughly $80B once you annualise the quarterly run-rate on page 14.

Check whether Amazon has published Q1 yet

Not yet — Amazon's FY26 Q1 filing isn't on EDGAR and no transcript exists. The index lists a 24 Apr expected date. I've used the Q4 call for the AWS commentary above; I'll flag it once Q1 lands.

Export that read-across as a markdown brief

Created a file, read a file

  • Write read-across brief for NVDAnvda_read_across.md
  • Presented file

Here's the brief — the three hyperscaler datapoints with the supply-constraint read at the top.

Nvda read acrossDocument · MD
  1. 1Tells you when a filing isn’t out yet
  2. 2Shows every step it took
  3. 3Hands you a brief you can download
  4. 4Pick your model: Claude, GPT or Grok

Choose the model
that answers

Claude, ChatGPT and Grok app icons side by side on a glass tray between dark rocks

Low, Medium, High or Max effort — regenerate any answer with a different model.

Read the source.Not the spin.

Start free trial

7-day free trial · Cancel anytime