UiPath Q2 FY2027: a $32M GAAP operating profit built on austerity

UiPath sells enterprise automation software — the robots and, increasingly, the AI agents and orchestration layer that let big companies run repetitive business processes without a human clicking through them. For most of its life as a public company the story has been the classic high-growth-software bind: fast revenue growth, persistent GAAP losses, and a stock-based-compensation bill so large that “profitability” existed only after you added it back. This quarter the shape changed. UiPath reported a $32 million GAAP operating profit against a $20 million operating loss in the same quarter a year ago, and a $89 million non-GAAP operating profit at a 22% margin.

The one-line finding: the profit is real, and it was manufactured by austerity, not growth. Revenue grew 13%, ARR 12%, and net new ARR was a thin $37 million. The operating-income swing came almost entirely from the cost side — operating expenses fell 6% year over year, research and development fell 15%, and stock-based compensation was nearly halved. The genuinely unusual footnote, and the one that inverts the standard critique of software names, is that UiPath’s share count is falling, not rising: buybacks over the first half outran dilution.

	Q2 FY27	Consensus
Revenue	$410.3M	~$397.9M	beat (~+3%)
Non-GAAP EPS	$0.15	$0.15	in line
GAAP operating income	+$31.6M	—	swing to profit
Net new ARR	$37M	—	thin
FY27 revenue guide	$1.789–1.794B	~$1.778B est.	raised above

The income statement, read two ways

Every figure below is from UiPath’s own statement of operations for the three months ended July 31, 2026, with the year-ago column from the same release.

Income statement	Q2 FY27	Q2 FY26	YoY
Licenses	$123.8M	$112.2M	+10%
Subscription services	$266.1M	$238.4M	+12%
Professional services & other	$20.3M	$11.2M	+82%
Total revenue	$410.3M	$361.7M	+13%
Gross profit	$329.7M	$297.3M	+11%
GAAP gross margin	80%	82%	−2 pts
Sales & marketing	$164.6M	$166.3M	−1%
Research & development	$83.4M	$98.3M	−15%
General & administrative	$50.1M	$52.9M	−5%
GAAP operating income (loss)	+$31.6M	−$20.2M	swing
GAAP operating margin	8%	−6%	+14 pts
Interest + other income, net	$21.3M	$23.5M	−10%
Provision for income taxes	$16.8M	$1.7M	—
GAAP net income	$36.1M	$1.6M	—
GAAP diluted EPS	$0.07	$0.00	—
Non-GAAP operating income	$89.0M	$62.3M	+43%
Non-GAAP operating margin	22%	17%	+5 pts
Non-GAAP net income	$80.9M	$80.3M	+1%
Non-GAAP diluted EPS	$0.15	$0.15	flat

Two facts jump out of that table and frame everything else.

First, the top line is decent but not fast. Total revenue of $410.3 million grew 13% and beat the Street’s ~$398 million. Subscription services — the recurring core — grew 12%. The 82% jump in professional services is off a tiny base ($11 million to $20 million) and is the lowest-margin line, so it flatters revenue growth while diluting gross margin, which slipped two points to 80%. Strip the professional-services noise and the durable, recurring business is growing at low-double-digits.

Second, the entire operating-income swing lives in the cost lines. Revenue rose $48.5 million year over year, but total operating expenses fell $19.5 million. R&D alone dropped $15 million (down 15%), sales and marketing edged down, and G&A fell. A company does not usually cut R&D 15% while claiming its future is a new AI product cycle; UiPath did, and that tension is the center of the story.

Did UiPath beat its own guidance in Q2 FY2027?

Yes. Revenue of $410.3 million came in above the top of the $395–400 million range UiPath guided to in its May 28, 2026 8-K, ARR of $1.938 billion was above the top of the $1.929–1.934 billion range, and non-GAAP operating income of $89.0 million was above the approximately $75 million guided. Year over year, revenue rose 13% and ARR 12%, as reported, and GAAP operating income swung to $31.6 million from a $20.2 million loss. UiPath also raised its full-year FY2027 revenue guide to $1.789–1.794 billion from the $1.776–1.781 billion it gave in May.

Where the profit actually came from: stock-based comp

The single largest driver of the swing is stock-based compensation. Here is the same cost, both years, straight from the reconciliation tables:

	Q2 FY27	Q2 FY26
Total stock-based compensation	$45.0M	$78.0M
SBC as % of revenue	11.0%	21.6%

Stock comp was nearly halved year over year — from $78 million to $45 million — and fell from 22% of revenue to 11%. That $33 million reduction is the single largest driver of the $52 million improvement in GAAP operating income — roughly two-thirds of the swing, with the rest of the P&L contributing the remainder. Because GAAP expenses include stock comp and non-GAAP excludes it, cutting SBC pulls the two profit figures toward each other: a year ago GAAP net income was a rounding-error $1.6 million while non-GAAP net income was $80.3 million, a gap of roughly $79 million; this quarter GAAP net income was $36.1 million against $80.9 million non-GAAP, a gap of about $45 million. The company did not manufacture the profit with an add-back — it manufactured it by handing out far less stock.

For most software companies the SBC line is the shareholder’s hidden bill: the reported non-GAAP profit is real cash, but the add-back quietly transfers ownership from existing holders to employees every quarter. UiPath is, unusually, running that dial in reverse — which is the next point, and the one worth understanding most.

The share count is going down, not up

This is the finding that separates UiPath’s quarter from the CrowdStrikes and Palantirs of the category, where 25%+ of revenue goes out as stock and the share count climbs.

• Weighted-average diluted shares fell to 523.0 million, from 542.9 million a year ago — down about 4%.
• Over the first half, UiPath repurchased $268.5 million of Class A stock (treasury stock on the balance sheet grew from $833.9 million to $1,092.8 million).
• Six-month buybacks of $269 million exceeded six-month non-GAAP adjusted free cash flow of $161 million — so the company is returning more than it generates, funded by drawing down its cash pile.

Put those together and UiPath this quarter is anti-dilutive: it issued less stock than it retired. That is a genuinely different profile from its high-growth software peers, and it is the strongest fact in the release for a per-share investor. The caveat is in the third bullet — buying back more than you generate is only sustainable while the balance sheet allows it, and UiPath’s cash and marketable securities fell from $1.689 billion at year-end to $1.405 billion, a $284 million draw over six months (buybacks plus a ~$149 million acquisition, partly offset by free cash flow). The buyback is a real return of capital; it is also spending the war chest.

The growth engine is soft, and that is the bear case’s foundation

The reason austerity — not growth — had to carry the profit is that the growth metrics decelerated.

• ARR of $1.938 billion grew 12%. UiPath manages the business on ARR, and 12% is slow for a company that still describes itself as riding an automation-and-AI wave.
• Net new ARR was $37 million. On a $1.9 billion base that is under 2% sequential expansion — the thinnest kind of growth quarter.
• Dollar-based net retention was 109%. Existing customers spent 9% more than a year ago. That is above 100% (net expansion, not contraction), but it is well down from the 115%–120% range UiPath posted in its faster years. Retention this soft means the installed base is barely expanding, and new-customer ARR is not filling the gap.

The bull answer is the product cycle: management is leaning the whole story on agentic AI and orchestration — the new Maestro Case and Maestro Flow products launched this quarter, pitched as the governed layer that coordinates AI agents, robots, and people across a business process. Founder-CEO Daniel Dines framed it directly: “AI is expanding what enterprises can automate, while increasing the need for the orchestration, governance, and exactness that deterministic automation provides.” That is a coherent thesis. It is also entirely a forward bet — none of it is visible yet in the 12% ARR growth or the $37 million of net new ARR.

Guidance: growth stays modest, margin does the work

UiPath raised the full-year revenue outlook, and the shape of the guide confirms the “profit through discipline” read.

Guidance	Q3 FY27	FY2027
Revenue	$440–445M	$1.789–1.794B
ARR (period-end)	$1.992–1.997B	$2.065–2.070B
Non-GAAP operating income	~$100M	~$445M

At the FY midpoints, revenue of about $1.79 billion implies roughly $963 million of second-half revenue against $829 million booked in the first half, and the ARR guide implies only about $130 million of net new ARR across the back half — still-modest growth. But non-GAAP operating income of ~$445 million on ~$1.79 billion of revenue is a ~25% margin for the year, up meaningfully. The FY revenue guide sits above the ~$1.778 billion analysts were carrying (per Benzinga). Management is telling you the plan: mid-teens-or-slower growth, margins expanding to do the heavy lifting.

The bear case, stated fairly

Not “the profit is fake” — it is real, GAAP, and above the operating line. The narrower version:

1. The profit swing is a cost story, not a growth story. Operating expenses fell while revenue rose; R&D was cut 15%. You can only cut your way to the first profitable comparison once, and cutting R&D during a product transition is the riskiest place to find the savings.
2. The recurring business is growing 12%. ARR up 12%, net new ARR of $37 million, and dollar-based net retention down to 109% describe a business whose expansion has slowed markedly. Software multiples assume growth; this is slow-growth software.
3. Non-GAAP EPS was flat year over year despite operating income up 43%. The improvement did not reach the per-share bottom line — a much larger tax provision ($16.8 million versus $1.7 million) absorbed it. Operating leverage that stops at the tax line does less for owners than it looks.
4. The buyback outruns free cash flow. Returning $269 million while generating $161 million draws the cash balance down; it supports the shrinking share count only as long as the balance sheet cooperates.

The bull case, from the same numbers

1. A genuine GAAP operating profit. Unlike many software peers whose GAAP operating line is still negative, UiPath’s operations made money on a GAAP basis — $32 million, before a cent of interest income. Interest and other income ($21 million) sit on top of an already-positive operating result, not underneath a loss.
2. The dilution problem is reversing. SBC at 11% of revenue and a share count down 4% make UiPath, right now, one of the few names in its category returning ownership to shareholders rather than diluting it.
3. Margins have real room. A 22% non-GAAP operating margin guided toward ~25% for the year, on a business that was losing money a year ago, is a fast structural improvement.
4. The AI-orchestration pivot is optionality, not yet cost. If Maestro converts the installed base’s interest in AI agents into net new ARR, the growth line reaccelerates on top of a cost base that has already been reset. The bet is unproven — but it is cheap to hold given the profitability is already there.

What to check at the Q3 report

Specific and falsifiable, in rough order of importance:

• Net new ARR and ARR growth. The whole question is whether growth reaccelerates or keeps sliding. The Q3 ARR guide of ~$1.995 billion implies only ~$57 million of net new ARR — watch whether even that modest bar is cleared, and whether the growth rate stabilizes above 10%.
• Dollar-based net retention. 109% is the softest it has been; another step down would say the installed base is stalling. A move back up would be the first hard evidence the AI pivot is landing.
• R&D spend. Whether the 15% cut holds or reverses. Sustained cuts fund the margin but starve the product; a reversal would say management is choosing the AI cycle over the near-term margin.
• Stock-based comp as a percent of revenue. Whether it stays near 11% — the number that makes the anti-dilution story durable.
• Buyback pace versus free cash flow, and the cash balance. Whether repurchases keep outrunning FCF and how far the $1.405 billion cash-and-securities pile is drawn down.
• Any first ARR contribution attributed to Maestro / agentic products. The only thing that turns the forward bet into a number.

The read

UiPath crossed a real line this quarter: a GAAP operating profit, a 22% non-GAAP operating margin, and — most unusually for its category — a shrinking share count backed by a live buyback. For a company that spent its public life losing money and diluting holders, that is a meaningful change in kind, not just degree, and it is the reason to take the quarter seriously.

The discipline is to see how it was done. This profit is the product of cutting costs — operating expenses down 6%, R&D down 15%, stock comp nearly halved — while the business grew only 13% on revenue and 12% on ARR, with net new ARR of a thin $37 million and retention easing to 109%. That is a company that has fixed its cost structure and now has to prove it can still grow. The agentic-AI orchestration pivot is the growth bet, and none of it is visible in the numbers yet. Watch two figures next quarter: net new ARR (does growth reaccelerate off $37 million, or keep fading?) and dollar-based net retention (does 109% stabilize and turn up?). The profitability is no longer the question. Durable growth is.

---

Sources and method. Company figures in this piece — total, licenses, subscription-services and professional-services revenue and their growth rates; GAAP and non-GAAP gross profit and margin; sales-and-marketing, R&D and G&A expenses; GAAP operating income of $31.6 million and the year-ago $20.2 million loss; non-GAAP operating income of $89.0 million and its 22% margin; interest and other income; the $16.8 million tax provision; GAAP net income of $36.1 million and $0.07 diluted EPS; non-GAAP net income of $80.9 million and $0.15 diluted EPS; stock-based compensation of $45.0 million; ARR of $1.938 billion, net new ARR of $37 million and the 109% dollar-based net retention rate; the balance-sheet cash and marketable securities of $1.405 billion, treasury-stock and share-count figures; six-month operating cash flow of $162.6 million, non-GAAP adjusted free cash flow of $160.8 million, $268.5 million of buybacks and ~$149 million of acquisition payments; and all Q3 and full-year FY2027 guidance — are UiPath’s own disclosures from its Q2 FY2027 results release (Form 8-K, Exhibit 99.1) filed 2026-09-03 for the quarter ended 2026-07-31, reported after the close. Prior-year comparatives are the year-ago columns in that same release. The Q2 FY2027 guidance (revenue of $395–400 million, ARR of $1.929–1.934 billion, non-GAAP operating income of approximately $75 million) and the earlier full-year FY2027 revenue guide of $1.776–1.781 billion are from the Q1 FY2027 results release (Form 8-K, Exhibit 99.1) filed 2026-05-28. Whether the quarter beat or missed its own guidance is judged only against that previously issued guidance and against the year-ago quarter. Not company figures, and attributed accordingly: the Zacks consensus (non-GAAP EPS of $0.15, in line, and the ~3.2% revenue beat implying a ~$398 million estimate) reported by Yahoo Finance, and the ~$1.778 billion full-year revenue analyst estimate reported by Benzinga. Our arithmetic: all year-over-year percentages, margins (e.g. GAAP operating margin of 7.7%, SBC at 11.0% of revenue), the share-count change, and the observation that six-month buybacks exceeded six-month free cash flow are computed from the disclosed dollars and are approximations rounded for readability; every table foots to the release. The reading that the profit was driven by cost discipline rather than growth, that the shrinking share count inverts the category’s usual dilution problem, and that the AI-orchestration pivot is an unproven forward bet, is our interpretation, not a company statement, and is not a forecast. This piece makes no claim about the share-price reaction.

Related: CrowdStrike Q2 FY2027, the security-software name where stock comp is still larger than the entire non-GAAP profit — the mirror image of UiPath’s shrinking-dilution quarter; Palantir Q2 2026, the other enterprise-software AI story where the gap between GAAP and non-GAAP profit is the whole debate; and Duolingo Q2 2026, a faster-growing software name to hold UiPath’s decelerating 12% against.

Tradune shows what you actually hold across every account, and whether a position has beaten a plain index fund since you bought it. Educational analysis, not investment advice.