3,000 to 11,800 in three quarters. The AI Receptionist nobody is talking about.
RingCentral's AI Receptionist (AIR) went from 3,000 paying customers in Q2 2025 to 11,800 in Q1 2026 — nearly 4x in three quarters. ACE grew +85% YoY to 5,200 customers. AI products are now ~10% of total ARR and growing rapidly. Operating margins exploded from 1.7% to 22.9% YoY as AI replaced internal headcount. Q1 2026 beat. Management raised full-year EPS guide. First-ever quarterly dividend declared. 4 brokers raised price targets post-earnings (Rosenblatt $50, Oppenheimer $50, Raymond James $55, Baird $45). No debt maturities until 2030. $581M FCF targeted for 2026. The legacy UCaaS bear case (Microsoft Teams kills RingCentral) is being quietly refuted by data. The market is anchored to 2023 sentiment. The 2026 print is a different company.
The market sees a dying UCaaS company. The data shows an agentic AI growth story.
RingCentral has spent the past four years buried under one of the most stubborn bear cases in software: Microsoft Teams is bundled "free" with every Office 365 license, so why pay for RingCentral? That narrative was right for the legacy seat-based business — and it remains the reason the stock trades at 8x earnings instead of 30x like other vertical SaaS names. But the bears are looking at the wrong revenue line. The new growth engine is agentic voice AI: AI Receptionist (AIR), AI Conversation Expert (ACE), AI Virtual Assistant (AVA), and now AIR Pro. These products grew from a standing start to nearly 10% of total ARR in 18 months. Customer counts are compounding 40%+ QoQ. Operating margins are up 12x YoY. And the company just paid its first-ever dividend. This is no longer a "dying SaaS" trade — it's an "agentic AI on a value multiple" trade.
Every business has phones nobody wants to answer. That is a $65B problem RingCentral now sells the solution to.
The fastest-growing customer-facing AI category in 2026 is not chatbots — it's voice agents. The reason is structural: most SMB and mid-market business interactions still happen by phone. Restaurants, dentists, law firms, contractors, healthcare clinics, real estate offices — they get tens of inbound calls per day, miss half of them after hours, and lose appointments, leads, and revenue every time. Until 2025, the alternative was hiring a receptionist or running a clunky IVR menu. Now, an AI voice agent — that sounds human, books appointments, routes calls, and updates the CRM — can be deployed in minutes. That is what AIR is. And 11,800 paying customers in 15 months from launch is how the market is responding.
AI Receptionist Customer Growth — The Hockey Stick
The Four-Product Agentic Voice Stack (RCAI)
RingCentral has organized its AI portfolio around the customer conversation lifecycle: before the call, during the call, and after the call. Each product targets a distinct buyer and price tier.
| Product | What It Does | Where It Sits |
|---|---|---|
| AIR (AI Receptionist) | Answers calls, books appointments, routes, updates CRM | BEFORE · SMB front desk · 11,800 customers |
| AIR Pro | Agentic voice platform · no-code Studio for custom agents | BEFORE · mid-market complex automation |
| AVA (AI Virtual Assistant) | Real-time agent assist · captures notes, surfaces recs | DURING · contact center agent productivity |
| ACE (AI Conversation Expert) | Post-call analytics, coaching, quality scoring | AFTER · supervisor/manager coaching · 5,200 customers (+85% YoY) |
The Killer Use Case: Replacing Missed Calls
RingCentral cites a Televero Health case study: deploying AIR contributed to a 97% patient satisfaction rate, a 14% monthly appointment increase, and over $200,000 in additional monthly revenue. Another, Keller Interiors (a Lowe's Home Improvement installation partner), deployed AIR to handle high call volumes across 33 locations without adding headcount. AIR is genuinely solving a problem business owners have been complaining about for decades — and at a price point ($30-100/month per business) that SMBs can authorize without procurement signoff. This is what product-market fit looks like.
Why This Refutes The Microsoft Teams Bear Case
The "Microsoft Teams kills RingCentral" thesis was about a specific battleground: employee-to-employee internal communications. RingCentral lost that battle. Microsoft won by bundling Teams with the Office 365 license customers already had. But AIR is a customer-facing product — it replaces the receptionist, not the employee chat tool. Microsoft does not have a comparable voice AI receptionist sold per-business. Google doesn't either. Zoom is building one but is behind. The category is wide open, RingCentral has the head start, and it's selling to the same customer base it already has the phone relationship with. The Teams bear case was correct for the legacy product. It's the wrong frame for the new one.
The Operating Leverage Story Nobody Is Modeling
The single most under-appreciated data point in Q1 2026: operating margin went from 1.7% to 22.9% YoY. That is not normal margin expansion. That is what happens when AI starts eating internal headcount. RingCentral has reduced stock-based compensation as a percentage of revenue by 420 basis points YoY, alongside operating expense discipline driven by AI-enabled automation across the support, sales, and engineering organizations. The company is now profitable in a way it wasn't in 2024, and the trajectory is structural, not cyclical.
Four brokers raised PTs on Q1. The reluctant bears are starting to capitulate.
A consensus rerate happens in phases. First, the Buy-rated bulls raise targets — easy. Then the Hold-rated skeptics raise targets (still Hold) — that's the signal. Then the Hold ratings flip to Buy. RingCentral is currently in phase two. Rosenblatt went $45→$50 (Buy). Oppenheimer $38→$50 (Outperform). Raymond James $40→$55 (Outperform). And critically — Baird raised $34→$45 while keeping Neutral. When a Neutral-rated analyst raises a PT by 32%, the rating is the lagging indicator. The next leg is the Neutrals flipping to Buy.
Tier 1 — Recent Analyst Actions (Post-Q1 2026)
| Firm | Action | Rating |
|---|---|---|
| Raymond James | PT $40 → $55 (+38%). "Q1 results in-line, strong AI traction." | Outperform |
| Oppenheimer (Timothy Horan) | PT $38 → $50 (+32%). Cited strong AI product adoption. | Outperform |
| Rosenblatt | PT $45 → $50 (+11%). "Beat on the key metrics." | Buy |
| Baird | PT $34 → $45 (+32%). Still Neutral — but Neutral analysts don't raise PTs 32% unless something changed. | Neutral |
| Trefis / Simply Wall St | Intrinsic value estimate ~68% upside. Sees re-rate based on AI ARR and FCF. | — |
| Consensus (~17 analysts) | Average PT ~$48 = +11% upside. Range $40-$55. Direction of every recent revision: UP. | Buy |
Tier 2 — Industry Recognition
The product-side validation has been quietly accumulating. RingCentral was named a Leader in the IDC MarketScape: Worldwide Communications Engagement Platforms 2026, called out specifically for its unified UCaaS, CCaaS, and CPaaS architecture and deep AI integration. Omdia Universe upgraded RingCentral from Challenger to Leader in its 2026 Customer Engagement Platforms report — the only category leader to advance two tiers YoY. Metrigy's 2026 MetriStar Report named RingCentral a Top CCaaS Provider — one of only three to earn the 2026 Top Provider Award for both CCaaS and WEM (Workforce Engagement Management). The product is winning awards specifically for the AI integration the stock is not yet being paid for.
Tier 3 — Customer Validation
| Partner / Customer | Action | Why It Matters |
|---|---|---|
| Cox Business | Launched "AI-First Contact Center Solution powered by RingCentral" — omnichannel platform for enterprise. | Tier-1 telco channel partner adopting RNG's AI stack |
| Spectrum Business (Charter) | RingCX + ACE rolled out to Spectrum Business customer base. Partnership expansion announced Q1 2026. | Cable channel distribution at scale |
| 11,800 AIR Customers | Healthcare (Televero Health · +14% appointments · +$200K/mo revenue), retail (Keller Interiors · 33 locations no headcount add), legal, hospitality, financial services. | SMB to mid-market across all major verticals |
| 5,200 ACE Customers | +85% YoY growth. Enterprise-grade conversation intelligence + coaching at scale. | Mid-market and enterprise CCaaS adoption |
| 1,700+ RingCX Customers | Native contact center crossing 1,700 — more than doubled YoY. >50% use AI features. | Cross-sell to existing UCaaS base is working |
The M&A Tail Risk (Or Catalyst)
A persistent Wall Street rumor: RingCentral has been mentioned as a potential acquisition target by larger telco players (T-Mobile, Verizon) looking to bolster their enterprise software stack, or by a private equity buyer attracted to the FCF profile. At $3.5B market cap, ~$580M annual FCF, no near-term debt maturities, and a category-leading AI product set, RNG is a clean strategic asset. This is not a base-case driver — speculating on takeover is not analysis. But the structural attractiveness as a takeout candidate is real, and acts as a downside cushion on the stock.
The phone company that learned to ship AI products. And finally turned profitable.
RingCentral, Inc. (NYSE: RNG). Founded 1999. Headquartered in Belmont, CA. Founder/CEO Vlad Shmunis still at the helm — a notable signal given how many UCaaS founders cashed out. The company started as a cloud-based business phone system competing with on-premise PBXes. Through the 2010s, it expanded into team messaging, video, and contact center to compete with Cisco, Zoom, Microsoft Teams. The bear case from 2021-2024 was that Teams would crush the seat-based UCaaS business. The bear case was partially right. The company responded by pivoting hard into AI-powered customer engagement starting in late 2024 — a strategy that is now visibly paying off in the operating metrics. ~83M shares out. Market cap ~$3.5B. No debt maturities until 2030.
Three Product Lines
| Line | What It Does | Strategic Role |
|---|---|---|
| RingEX (UCaaS) | Unified comms: phone, video, messaging | Legacy core · seat-based · stable but commoditizing |
| RingCX (CCaaS) | Native AI-first contact center | Mid-market growth · 1,700+ customers · cross-sell engine |
| RingCentral AI (RCAI) | AIR · AIR Pro · AVA · ACE · RingSense | The growth engine. ~10% of ARR, compounding fast |
The Strategic Architecture — "Swiss Army Knife"
RingCentral organizes its commercial offering around customer size, with progressively more capability at each tier:
| Segment | Solution | AI Component |
|---|---|---|
| SMB | RingEX + AIR + Customer Engagement Bundle (CEB) | AIR handles ~90% of inbound call routing |
| Mid-market | RingCX + AIR Pro + AVA + ACE | Full agentic stack with no-code Studio |
| Enterprise | RingCentral Contact Center (with NICE) + RCAI | Enterprise CCaaS via NICE partnership |
How AIR Pro Works (The Newest Product)
AIR Pro is a voice-first, omnichannel AI agent platform announced March 2026. The differentiator: AIR Pro Studio, a no-code environment where any non-technical user can design, build, and deploy custom voice and digital AI agents in minutes using natural language. The competitor reference points are Google Dialogflow, Amazon Lex, OpenAI's GPT Builder — but built specifically for business voice with enterprise governance, telephony integration, and the RingCentral cloud number/PBX layer underneath. This moves RingCentral from a "buy AIR off the shelf" SKU to a "build your own AI agents on our voice platform" platform play. The TAM expands materially.
Recent Strategic Moves
Feb 2026: AIR international expansion (UK, Australia, Spanish, French language support). Feb 19, 2026: Q4 2025 earnings — AIR crosses 8,300 customers. Feb 2026: AIR Everywhere launched — works with any third-party phone system, expanding TAM beyond RingEX customers. March 2026: AIR Pro launched with no-code Studio. March 2026: $609M convertibles paid off at maturity — no debt maturities until 2030. May 7, 2026: Q1 2026 earnings — beat, raised, first-ever dividend declared ($0.075 quarterly), AIR crosses 11,800 customers. May 2026: Customer Engagement Bundle for Microsoft Teams announced — bringing lightweight CCaaS into Teams environment.
Operating margin 22.9%. Up from 1.7% YoY. That is not normal.
Q1 FY26 Results (Reported May 7)
| Metric | Q1 FY26 | Δ YoY / Note |
|---|---|---|
| Revenue | $644.2M (beat $642.7M est) | +5.3% YoY · in line |
| Adjusted EPS | $1.20 (beat $1.17 est) | +2.7% beat |
| Adjusted Operating Income | $147.3M | 22.9% margin · 0.8% beat |
| Operating Margin | 7.8% GAAP / 22.9% adj | UP from 1.7% GAAP YoY |
| Net Income | $30.6M | Swung from loss YoY |
| Annual Recurring Revenue | $2.71B | +7% YoY |
| Billings | $632.4M | +5.6% YoY |
| FY26 Adjusted EPS Guide | $4.93 midpoint | RAISED +1.3% |
| FY26 Revenue Guide | $2.63B midpoint | In line with consensus |
The AI Customer Metric That Matters
| Product | Customers | Growth |
|---|---|---|
| AIR (AI Receptionist) | 11,800 | +40% QoQ · +293% in 9 months |
| ACE (Conversation Expert) | 5,200 | +85% YoY |
| RingCX (CCaaS) | >1,700 | More than 2x YoY · >50% use AI |
| Customer Engagement Bundle | >5,000 | ~40% attach rate for paid AI features |
The Valuation Math
At $43.16 spot and ~$3.5B market cap on $4.93 FY26 EPS, RNG trades at ~8.7x forward EPS. The US Software industry average P/E is 27x. The UCaaS peer average is ~46x. RNG trades at less than one-third the multiple of its software peers despite faster operating margin expansion, growing AI ARR, and a fresh dividend. Simply Wall St's discounted cash flow estimate puts intrinsic value ~68% above current price. The Trefis intrinsic value framework points to similar upside. The valuation gap is the entire trade. The bears say it's deserved because of Microsoft Teams seat compression. The data says the AI growth offsets seat compression with room to spare.
Capital Return Profile (NEW)
| Item | 2025 | 2026 |
|---|---|---|
| Quarterly Dividend | $0 (no dividend) | $0.075 (NEW · first ever) |
| Forward Dividend Yield | — | ~0.7% (small but symbolic) |
| Q1 Buyback | — | $81M (2.6M shares) |
| Buyback Authorization Remaining | — | $418M (~12% of cap) |
| Debt Maturity Wall | $609M (March 2026) | Cleared · nothing due until 2030 |
| 2026 FCF Target | — | ~$581M · ~16% FCF yield |
The capital return story is the cleanest signal a software company can send. RingCentral has formally crossed the line from "survival" to "compounder." First-ever dividend says management has high enough confidence in FCF durability to commit to a recurring payout. That's a different stock than the one that traded sub-$25 in 2024.
Analyst Coverage
17 analysts. Average rating Buy. Average PT ~$48 (+11% from spot). High PT $55 (Raymond James). Most recent revisions all UP. The 32% PT hike from Neutral-rated Baird is the analyst tell of the year — when a Hold analyst raises a target by a third, the rating is mechanically behind the data. Watch for Baird, JPMorgan, and other Neutrals to flip Buy over the next 1-2 quarters. That is when the consensus rerate completes.
Trade ideas like this, before they hit the timeline.
Join Discord →Lost the Teams war. Winning the AI Receptionist war nobody else is contesting.
RingCentral's competitive landscape splits cleanly into two battles. The legacy battle — UCaaS seat-based communications — is essentially over, and RNG lost. Microsoft Teams won by bundling with the enterprise license customers already had. But the new battle — agentic voice AI for customer engagement — is a category Microsoft, Google, and Zoom have not yet entered in a meaningful way. RingCentral has a genuine head start and is building a moat customers can't easily switch out of.
The Three Battles RNG Is Fighting
| Battle | Status | What's At Stake |
|---|---|---|
| UCaaS Seats (vs MSFT Teams) | Lost · seat-based business commoditizing | Legacy revenue line · structurally declining |
| CCaaS (vs Zoom, Five9, NICE, Genesys) | Winning share · RingCX > 1,700 customers | Mid-market growth · 50%+ already use AI |
| Agentic Voice AI (vs no one) | Leading · AIR 11,800 customers · 40% QoQ growth | Defines next 5 years of the business |
The Three Competitive Dynamics That Matter
RNG vs Microsoft Teams. The smartest strategic move RingCentral has made in five years: stopped fighting Teams and started integrating with it. "RingCentral for Microsoft Teams 2.0" lets enterprise customers use RingCentral's superior telephony inside the Teams interface. The new Customer Engagement Bundle for Microsoft Teams announced May 2026 brings lightweight contact center capabilities into the Teams environment — explicitly positioning RNG as the AI agent layer on top of Microsoft's collaboration product. Microsoft sells the rails. RingCentral sells the AI agents that run on top. This is the strategic flip from competing on collaboration UX to competing on AI workflows. RNG vs Zoom (ZM). Zoom pivoted from video to CCaaS in 2024, launching Zoom Contact Center. The product is competent but lacks RingCentral's vertical-specific AI features (healthcare, hospitality, construction) and lacks the breadth of agentic voice products. ZM trades at ~20x EPS vs RNG at ~8x — that 2.5x gap is the multiple-expansion thesis on RNG. RNG vs Five9 (FIVN). Five9 is the pure-play CCaaS comp. Trades at a higher multiple but has slower growth and fewer agentic AI products. Recently bought by ServiceNow rumors swirled in 2024 (didn't happen). FIVN's existence suggests the CCaaS category is being valued reasonably — RNG should re-rate toward FIVN's multiple as its CCaaS revenue grows.
The "Swiss Army Knife" Advantage
RingCentral's argument to customers — increasingly accepted by analysts — is that buying multiple AI vendors creates integration nightmares. One vendor handles UCaaS, CCaaS, AI receptionist, agent assist, conversation analytics, workforce engagement. That bundling is what won the IDC and Omdia Leader rankings. For SMBs in particular, the appeal of one bill, one login, one support contact is real. Salesforce Service Cloud + Twilio Voice + a separate AI agent vendor is an integration project. RingCentral is a checkbox.
The Channel Distribution Edge
RingCentral's distribution through Cox Business, Spectrum Business (Charter), AT&T, and other carrier partners is structurally underrated. These channels carry RNG's products to tens of thousands of SMBs that would never be sold to directly by RNG's enterprise sales team. Spectrum Business now offers RingCX + ACE to its customer base. Cox Business launched its AI-first contact center solution "powered by RingCentral." This is the BlackBerry/QNX-style royalty model in customer engagement — channel partners do the sales work, RNG collects the recurring revenue.
The bull case is now operational data. The bear case is a multi-year scar.
Bull Case
- AIR customer growth 3K → 11,800 in 9 months (+293%). 40% sequential QoQ. The killer datapoint nobody is modeling.
- Operating margin 22.9% (up from 1.7% YoY). AI replacing internal cost base. Structural, not cyclical.
- FY26 EPS guidance RAISED. $4.93 midpoint vs prior. Beat-and-raise pattern intact.
- First-ever quarterly dividend declared ($0.075). Symbolic but real — company crossed line from survival to compounder.
- $418M buyback authorization (12% of market cap). $81M repurchased in Q1 at sub-$45 levels.
- $609M of convertibles paid at maturity. No debt maturities until 2030. Balance sheet finally clean.
- ~$581M 2026 FCF target. ~16% FCF yield on current market cap.
- 4 brokers raised PTs post-Q1. Including Baird Neutral raising 32% — the analyst tell.
- Cox Business + Spectrum Business + Microsoft Teams all channel-distributing the RNG AI stack.
- ~8x forward P/E vs ~46x UCaaS peer average. The multiple-expansion case is mathematical.
Bear Case
- Microsoft Teams is bundled "free" with Office 365. Legacy UCaaS seat-based business commoditizing for the foreseeable future.
- Revenue growth only +5.3% YoY in Q1. Not the 20%+ growth that gets full software multiples.
- AI cannibalization risk. If AIR is so effective that customers need fewer human agents, seat-based revenue declines faster than AI ARR grows.
- Multi-year stock scar. Down from $400+ peak in 2021 to $43 today. Burned cohorts of investors. Sentiment takes time to mend.
- AIR customer counts are misleading. 11,800 customers — but at what average revenue per customer? Could be very low, meaning ARR contribution is small.
- Vlad Shmunis still CEO. Founder-led can be a strength or a stagnation flag. The market sometimes sees founder retention as resistance to change.
- Zoom, Google, and AWS could enter agentic voice AI. The 18-month head start is real but not insurmountable. Anyone with an LLM can build "AI receptionist."
- P/E premium vs broader software (8x vs 27x industry). Cheap relative to UCaaS peers but actually premium vs the broad software universe. Not unambiguously a bargain.
- SBC still ~10% of revenue. Improving but high. Dilution risk if margin gains slow.
- Negative market reaction to Q1 beat (stock fell 6%). Sentiment is fragile. A miss on Q2 could reset the entire narrative.
Even the cautious case is +27%. The full re-rate is a double.
Position Structure
RNG is a core software-AI long, not a speculative bet. The setup is structurally similar to a beaten-down growth name where the operating data has already confirmed the thesis but the multiple hasn't caught up. 3-5% common stock position is appropriate sizing — same weight class as the other AI infrastructure names in the book. The structure should be primarily equity with optional Jan 2027 calls as a defined-risk overlay.
The Three Payoff Vectors
This trade has three distinct ways to win that don't require all to fire:
(1) Earnings power. $4.93 FY26 EPS guide. Beat-and-raise pattern intact. Each beat compounds the EPS base for forward valuation. Even with no multiple expansion, FY27 EPS at $5.50+ on an 8x multiple is $44 — flat to current. With moderate multiple expansion to 12x, that's $66.
(2) Multiple re-rate. 8x forward EPS is too cheap for a software company growing operating margin 12x YoY with new product traction. Re-rating to even half the UCaaS peer multiple (23x vs 46x average) puts the stock at $113. This is the largest single source of upside.
(3) Capital return acceleration. $418M buyback authorization + $0.075 dividend + ~$581M annual FCF. The company is in position to retire 10%+ of float per year while still raising the dividend. Compounding effect on per-share metrics matters over multi-year horizons.
Options Considerations
RNG implied volatility is moderate (~40-50% on most expiries). Liquidity is reasonable but not great — bid-ask spreads can widen on far-dated strikes. Jan 2027 $50C or $55C provide leverage on the base/bull case with manageable theta given the 8-month runway through Q3 2026 earnings. For a multi-quarter inflection thesis, equity is the primary tool — options are an optional overlay, not the trade structure. Sized appropriately, a small allocation to ATM-to-slight-OTM LEAPs adds asymmetric upside without compromising the equity sleeve.
Catalyst Calendar
| Catalyst | Window | Impact |
|---|---|---|
| Dividend Ex-Date | Jun 2, 2026 | First-ever dividend payable Jun 11 |
| Q2 2026 Earnings | Early August 2026 | The decisive print. AIR customer count + operating margin trajectory. |
| AIR Pro Customer Traction | Q2-Q3 2026 | First quarters of Pro adoption · upsell rates |
| Customer Engagement Bundle for Teams | 2H 2026 | Distribution into Microsoft customer base |
| Q3 2026 Earnings | November 2026 | First full quarter of AIR Pro at scale |
| FY27 Guidance | February 2027 | Cements the trajectory for the rerate |
| Potential M&A Activity | Anytime | Persistent telco / PE rumor · downside cushion |
Cheapest by multiple. Highest by FCF yield. Only one with hockey-stick AI customer growth.
| Ticker | Mkt Cap | Fwd P/E | Rev Growth | FCF Yield | AI Hook |
|---|---|---|---|---|---|
| RNG | $3.5B | ~8.7x | +5.3% | ~16% | AIR · ACE · AVA · agentic voice |
| ZM | ~$24B | ~18x | +3% | ~14% | Zoom Contact Center · AI Companion |
| FIVN | ~$2.2B | ~14x | +8% | ~9% | Pure-play CCaaS · Genius AI |
| NICE | ~$10B | ~12x | +9% | ~10% | Enterprise CCaaS · CXone Mpower |
| TWLO | ~$15B | ~24x | +10% | ~7% | CPaaS · Twilio Voice + Verify |
| MSFT | ~$3.5T | ~32x | +13% | ~2.5% | Teams bundle + Copilot |
The Three Comps That Matter
RNG vs ZM: Both UCaaS-origin companies pivoting to AI-driven CCaaS. ZM at 18x EPS vs RNG at 8.7x — RNG is roughly half the multiple. Both have decelerated to mid-single-digit growth. ZM has stronger video brand recognition; RNG has stronger AI receptionist product traction. RNG re-rating to even ZM's 18x P/E on $4.93 EPS = $89 stock. That's the math the bulls are running. RNG vs FIVN: Direct CCaaS pure-play comp. Similar market cap ($3.5B vs $2.2B). RNG slightly cheaper P/E with higher FCF yield. Both have AI products, but RNG's AIR customer counts are growing faster than anything Five9 has disclosed. If FIVN deserves 14x, RNG deserves 14x+. RNG vs NICE: NICE is the enterprise CCaaS leader. Trades at 12x with similar FCF yield. NICE is actually RNG's enterprise channel partner — they sell together to the largest customers. RNG benefits from NICE's enterprise validation while having faster SMB growth.
Long RNG. Core position. Equity primary, LEAPs optional.
RNG is one of those rare setups where the operating metrics have decisively moved before the multiple. AIR went from 3,000 to 11,800 customers in 9 months. Operating margins jumped from 1.7% to 22.9% YoY. The company declared its first-ever dividend, paid off the convertibles maturity wall, and four brokers raised price targets post-Q1. The bear case — Microsoft Teams kills RingCentral — was right about the old business and is wrong about the new one. The new business is agentic voice AI for SMB and mid-market customer engagement, and it has product-market fit Microsoft has not yet contested. Trading at 8.7x forward EPS with 16% FCF yield and a 22.9% operating margin run-rate, this is the cheapest stock in the entire customer engagement category. The multiple-expansion thesis is mathematical, the catalyst is Q2 earnings in August, and the downside cushion is the persistent M&A rumor. Size as a core software-AI long.