Should You Forget Palantir and Buy These 2 Under-the-Radar AI Stocks Instead?

Palantir is praised for accelerating revenue but is very expensive, trading at a forward P/S of 47, limiting near-term upside. UiPath is transitioning from RPA to an agentic AI orchestration platform (Maestro), showing ARR growth acceleration and trading at a forward P/S of 3 and forward P/E of 13, positioning it as a cheap growth play. ServiceNow, with AI Control Tower and security acquisitions Armis and Veza, trades below a forward P/S of 6.5 and forward P/E under 24 while growing revenue ~20%, making it another attractive AI-orchestration candidate.

Analysis

Enterprise orchestration of agentic AI is shaping into a two-layer contest: workflow incumbents that already own processes and identity will win share of wallet faster than best-of-breed agent startups. The second-order beneficiary set includes observability, identity/permissions, and edge-inference vendors because orchestration multiplies the surface area that needs governance and low-latency execution. Expect M&A pressure inside large SaaS buyers to package orchestration + security rather than buy point solutions over the next 12–24 months.

Hardware and cloud economics are an underappreciated lever: as orchestration directs many short-lived LLM calls, cloud vendors and GPU suppliers capture a larger fraction of total solution economics, squeezing software gross margins unless vendors can productize predictable metered pricing. This creates a divergence where software companies that (a) can offload inference costs to customers or (b) offer compression/quantization toolchains will preserve margins; others will see accelerating, sticky revenue but lower incremental gross margins. Look for deal structures (consumption passes, hybrid on-prem appliances) to proliferate as customers hunt predictable TCO.

Catalysts and risks cluster by timeframe. In the next 3–9 months, contract pilots and early reference wins can drive re-rating for platform players; over 12–36 months, security incidents, negative TCO studies, or a sustained macro capex pullback can reverse momentum. The consensus is too binary: agent orchestration is not an all-or-nothing replacement of RPA or ITSM — it is a capital-light layer that will reallocate value across software, cloud, and hardware, creating both consolidation winners and margin casualties.

Key Decisions for Investors

• Long PATH (12 months): establish a 2–3% portfolio position in PATH equity or buy a 9–12 month 1:2 call spread to cap premium. Thesis: governance moat + fastest transition to orchestration; target +30–40% upside, stop at -18% if enterprise ARR growth stalls.
• Long NOW (12–24 months): accumulate a 1.5–2% position or buy a 12–18 month call calendar to play multi-year workflow integration optionality. Target +35–45% with downside protection via a 20% trailing stop; reward accrues from cross-sell and security-integration synergies.
• Pair trade (6–12 months): long NOW / short PLTR dollar-neutral (equal notional). Rationale: hedge macro/software multiple moves; target capture of 15–25% relative outperformance if market reprices premium AI winners. Risk: PLTR executes and multiple expands—limit pair to 1–1.5% net exposure and cut at 12% adverse spread move.
• Small asymmetric NVDA play (6–12 months): buy a limited-size call spread to express higher sustained inference demand while capping premium outlay. Expect material upside if orchestration adoption drives incremental GPU cycles; size at <1% portfolio to control event risk versus potential 2–3x upside on the spread.