Earnings season used to be a massive period of stress. On the busiest nights, 15 companies would report, of which we might have 5 positions, within a one-hour window. Every analyst knew the drill: the press release hit, you scrambled to extract the important numbers, you cross-checked against guidance from last quarter, you tried to figure out what was actually new versus what was noise, and by the time you had a coherent view, the stock had already made its move and your PM and traders had already asked you multiple times what you thought and why the stock was up or down 5%.
This quarter I tried Daloopa’s ‘Earnings Flash’ skill with Oracle reporting its fiscal Q3 earnings. The Earnings Flash is one of Daloopa’s pre-built skills — specialized instructions that run via the MCP for end-to-end analyst workflows. Within roughly a minute of the release, I had a structured earnings flash built on Daloopa’s verified data: verdicts, a key numbers table with every figure cited back to the source filing, guidance versus actuals, risk flags, and read-throughs to adjacent names. Not a finished thesis. A first read. But a first read with real numbers, which is exactly what you need at 4:05 pm when everything is reporting at once.
What the flash caught immediately
The headline verdicts wrote themselves once the data was in front of me:
The beat was real and broad. Total revenue came in at $17.2 billion, up 22% YoY in USD, and Oracle flagged it as the first quarter in over 15 years where organic revenue and non-GAAP EPS both grew 20% or more. Non-GAAP EPS of $1.79 beat guidance in both USD and constant currency. Total cloud revenue hit $8.9 billion, at the high end of constant currency guidance and above USD guidance.
Infrastructure is the entire story. Cloud Infrastructure revenue of $4.9 billion grew 63% YoY versus the prior year’s $3.0 billion. That is the AI demand signal the market has been focused on, and the flash surfaced it in seconds because the IaaS and SaaS series were already split out in Daloopa’s data; no manual table parsing required.
The guidance raise was real. Cloud revenue growth guidance was raised to 46% to 50% in USD, up from a prior 40% to 44% range. RPO of $553 billion, up 325% YoY, is worth looking into further, and the flash framed it exactly that way.
What the flash surfaced beyond the headline
This is where speed plus structure beats speed alone. A fast headline scrape tells you “beat and raise.” The Earnings Flash skill Daloopa built, on top of their clean fundamentals, tells you where to dig:
SaaS is decelerating while IaaS accelerates. Cloud Applications revenue of $4.0 billion grew only 8% YoY. That divergence is worth tracking — whether Oracle’s cloud story broadens beyond OCI consumption deals will be a key forward question. If Fusion and NetSuite stay in single digits, the ERP share narrative will be a focal point for follow-on analysis. It is a data point that fits into the broader conversation around AI’s impact on traditional software growth.
The path to operating leverage. Operating income of $5.5 billion grew roughly 7% YoY against 22% revenue growth, with depreciation increasing to $2.2 billion from $1.7 billion in the prior quarter. The capex bill for the AI buildout is landing in the P&L, and margin expansion has not yet followed.
The GAAP versus non-GAAP gap is wide. GAAP EPS of $1.27 against non-GAAP of $1.79 is a roughly 40% spread, driven by amortization and stock comp add-backs. Worth tracking every quarter as a quality of earnings item.
Q4 EPS guidance came down. New guidance of $1.92 to $1.96 sits below the prior range despite the Q3 beat. On a fast night, that is precisely the kind of detail that gets missed, and it is the kind of detail that explains why a “beat and raise” stock can trade down anyway.
Why this matters on busy earnings days
The value here is not that AI wrote a report. The value is the compression of the worst part of an analyst’s job — the mechanical work of pulling four quarters of comparables, lining actuals against guidance with the correct one-quarter offset, and splitting segments correctly used to consume the first 30 to 45 minutes after every release. That is exactly the window when the information is most valuable, and the analyst is most overloaded.
With Daloopa’s MCP, the Earnings Flash skill pulled the company, resolved the latest quarter from the data itself rather than guessing from the calendar, fetched the fundamentals with source-linked IDs, searched the actual 8-K and transcript, and grabbed the stock price for context. Every fundamental number in the output links back to the exact place in the filing it came from, so when my PM asks, “Where is that 63% IaaS number from?” the answer is one click, not a frantic ctrl-F through a press release.
The flash does not replace the deep dive. It replaces the scramble before the deep dive. On a night with five of your names reporting, that is the difference between having a view on all five and having a view on two.
The analysis is still yours. The hours you used to spend getting to the starting line are not.
