Soundhound.com

SoundHound AI completed its acquisition of LivePerson on September 4 — a deal the market has consistently described with a $43 million headline price tag that understates the real financial commitment by a factor of seven. The actual cost of the transaction, once SoundHound's concurrent retirement of LivePerson's outstanding secured debt is counted, was approximately $304 million: $42.8 million in cash and stock to LivePerson's equity holders, plus $261.2 million in cash and stock to retire LivePerson's First Lien and Second Lien Convertible Secured Notes. That full figure is what SoundHound spent to bring LivePerson's 1 billion monthly customer messages and its roster of 25 Fortune 100 clients into the combined company's orbit, as detailed in SoundHound's Q2 2026 10-Q filing.

The deal is the fifth acquisition in SoundHound's strategic expansion since going public in 2022, and the most consequential. It merges two distinct architectures — SoundHound's OASYS voice and agentic AI platform with LivePerson's Conversational Cloud text and messaging infrastructure — into what the combined company is billing as the first natively omnichannel enterprise AI solution: one system capable of handling customer interactions across phone calls, web chat, SMS, social messaging, and in-app channels without fragmenting the underlying AI or requiring customers to repeat themselves when switching channels. A September 4 press release confirmed the deal close and described functional integration as already underway.

What OASYS Actually Does — and Why the LivePerson Data Matters

To understand why enterprise buyers might care about this merger, it helps to understand what distinguishes OASYS from a standard AI chatbot or voice agent. Most enterprise conversational AI platforms require engineering teams to build individual agents — manually defining workflows, training on domain data, and maintaining the system as business processes change. OASYS, which SoundHound launched on May 5, 2026, operates differently: it manages the full lifecycle of an AI agent fleet autonomously. The system creates new agents, orchestrates them during live interactions (dynamically selecting and coordinating multiple specialized sub-agents within a single customer conversation), evaluates their performance, identifies gaps, and engineers its own updates — which human experts then review before deployment.

The practical effect, if it works as claimed, is that a bank deploying OASYS to handle customer service calls gets an AI system that improves itself based on call outcomes without continuous developer intervention. SoundHound's patented Human Assisted Resolution (HAR) technology serves as the guardrail: for cases the AI cannot handle confidently, it routes to human experts rather than failing silently. SoundHound was named winner of the 2026 AI Breakthrough Award for agentic AI company of the year, with OASYS cited as the defining platform.

Where LivePerson enters is in the training data and the channel infrastructure. LivePerson's Conversational Cloud has been processing enterprise customer messages since the company pioneered webchat for brands in the late 1990s. Its Intent Manager — the real-time natural language understanding engine at the core of the platform — is grounded in over 20 years of proprietary data and more than one billion messaging transcripts across financial services, retail, healthcare, airlines, and telecommunications. That labeled intent data — what customers actually want from banks, what queries airline passengers most frequently escalate, what healthcare scheduling questions resist automation — is the kind of domain-specific training asset that is expensive and slow to replicate. SoundHound's voice platform processes billions of voice interactions annually. Combining those two datasets gives the merged entity tens of billions of annual customer interactions across both modalities, which is expected to improve the accuracy and containment rates of the combined OASYS-powered platform. The April 2026 deal announcement outlined the combined enterprise value opportunity at approximately $500 million from existing customers alone.

Does the Omnichannel Claim Hold Up?

The strategic logic is straightforward: SoundHound had voice, and no serious digital messaging infrastructure. LivePerson had digital messaging at scale, and no serious voice AI. The gap between them maps directly onto the gap most enterprises experience when they try to deliver a consistent customer service experience — a customer who calls the helpline gets a different AI experience than one who opens the web chat, even if both are asking the same question. Session context is lost between channels. The AI has to start over.

What the combined OASYS + Conversational Cloud platform claims to solve is unified session identity across channels. A customer who opens a billing dispute by phone can be handed to web chat and then to SMS — with the AI retaining the full context of the prior conversation, because both voice and text are processed through the same OASYS orchestration layer rather than separate systems. Whether that works in production at enterprise scale is the integration question SoundHound's engineering teams are now responsible for answering.

The context preservation claim is technically non-trivial. Voice and messaging differ significantly in latency tolerances (voice requires near-instantaneous response; messaging tolerates seconds), in the nature of input (voice requires real-time speech-to-text conversion before NLU can run; messaging operates on already-text input), and in the session model (a phone call has a defined start and end; a messaging thread may span days). Building a unified orchestration layer that handles both without degradation is a meaningful engineering challenge, and one the merged company has not yet demonstrated in a production cross-channel handoff scenario. The company says combined product capabilities will reach clients in the coming quarters.

Real Financial Picture: Growth With Significant Losses on Both Sides

SoundHound itself is not yet profitable. The company posted record Q2 2026 revenue of $61.9 million — up 45% year-over-year and the largest quarter in the company's history, with the Americas segment growing 85% year-over-year — but recorded a GAAP net loss of $42.8 million for the quarter. The adjusted EBITDA loss for Q2 was $9.6 million.

Full-year 2026 revenue guidance (not yet reflecting LivePerson's post-close contribution) stands at $230 million to $260 million. The company is guiding investors toward a 2027 revenue range of $350 million to $400 million post-merger, with LivePerson expected to contribute approximately $100 million of that total. SoundHound's 2026 Q2 guidance explicitly excluded the LivePerson contribution; updated guidance reflecting the acquisition close is expected before year-end.

LivePerson's own recent financials compound the picture. The company reported a net loss of $134.2 million in 2024 and approximately $67.2 million in 2025, and its stock had declined approximately 100% from its 2021 peak by the time the acquisition closed. That history is precisely why SoundHound could acquire the company for what amounts to a fraction of its peak valuation — and precisely why analysts flagged integration risk as the central question when the deal was announced in April.

The company's response to the profitability question is John Collins, the newly appointed CFO. Collins served previously as CFO, COO, and Interim CEO at LivePerson and has a specific track record relevant to the integration task ahead: he led LivePerson's operational transformation from more than $100 million in annual cash burn to positive free cash flow within a single fiscal year, executed cost-reduction programs exceeding $200 million, and managed multi-year debt restructurings that captured $227 million in debt discount. His background is outlined in the September 4 deal-close announcement. His mandate is explicit: drive margin expansion, enforce cost discipline, and demonstrate that the combined company can convert strong top-line growth into sustainable profitability. Company management has projected a path to profitable growth by early 2027, while many analysts do not expect sustainable profitability until 2028.

What Enterprises Are Actually Buying Here (and Who Else Wants Their Budget)

The combined SoundHound-LivePerson entity now competes directly in one of the most actively contested markets in enterprise technology. Gartner's February 2026 forecast for agentic AI software spending projected that enterprise spending would reach $985 billion by 2030, growing at a 62.7% compound annual growth rate from 2025 — a number that has every major AI platform vendor scrambling for footprint before the market matures.

The competitive landscape the combined company enters is formidable. Salesforce unveiled its Agentforce Contact Center earlier in 2026 — positioning the product as an AI, omnichannel, and CRM solution in one — and dominated Enterprise Connect 2026 with the announcement, per Aragon Research. Salesforce's advantage is its CRM footprint: for any enterprise where Salesforce is already the system of record for customer data, Agentforce has native access to customer history that a vendor without that CRM integration has to replicate through additional engineering work.

Genesys acquired agentic workflow automation company Pinkfish in June 2026, adding more than 500 integrations and 25,000 MCP tools to its Genesys Cloud platform — another consolidation play aimed at solving the same fragmentation problem SoundHound is targeting. NICE acquired Cognigy for approximately $955 million in September 2025. Amazon Web Services and Microsoft are present throughout the enterprise contact center stack through Amazon Connect, Lex, and Copilot Studio. The contact center AI sector is consolidating rapidly, and the SoundHound-LivePerson combination is the latest move in that consolidation race.

The differentiated bet SoundHound is making is that the enterprise AI contact center market will ultimately reward independence from a CRM vendor and natively trained voice capability. Salesforce's Agentforce is strongest where Salesforce CRM is already deployed; SoundHound's claim is that OASYS can serve any enterprise regardless of CRM, and do so across both voice and digital natively. The market will determine which architectural bet proves correct. But there is now a non-trivial alternative to the CRM-anchored approach — backed by 25 Fortune 100 relationships and 750+ patents — competing for the same enterprise technology budget.

Is the 2027 Revenue Target Realistic?

The most specific question SoundHound needs to answer is whether LivePerson's existing enterprise relationships survive the transition. D.A. Davidson analyst Gil Luria asked directly on SoundHound's Q1 2026 earnings call about expected revenue attrition from LivePerson's customer base during integration. COO Mike Zagorsek confirmed that SoundHound models customer churn conservatively in its integration plan, targeting a "transformational turnaround" — a new revenue baseline before growth resumes. That baseline — LivePerson's retained and migrated revenue — is the foundation beneath the $100 million LivePerson contribution embedded in the 2027 guidance.

Enterprise AI platform migrations typically take 12 to 24 months, with engineering work across routing, conversation flows, reporting, and CRM integrations. Enterprises running non-SoundHound infrastructure face a different integration model than those on platforms already connectorized to the Conversational Cloud. The customers that matter most near-term, per SoundHound CEO Keyvan Mohajer, are large enterprise renewals — he cited a Fortune 100 insurance client as the first-wave target, with all new customers expected to onboard to OASYS immediately.

How Should Enterprise Technology Buyers Think About This?

For an enterprise technology buyer evaluating customer engagement platforms in the second half of 2026, the SoundHound-LivePerson combination is now a legitimately new category: a purpose-built, independent, omnichannel AI platform with no legacy CRM dependency and native voice capability backed by over two decades of enterprise messaging data. It is the first credible alternative for enterprises that want unified voice and digital AI without committing to a CRM vendor's ecosystem.

The caveats are real: neither company has demonstrated sustained profitability, the integration of two complex enterprise platforms is a multi-year engineering project still in its earliest stages, and the competitive pressure from Salesforce, Genesys, and Amazon means the window for execution is not unlimited. Enterprise buyers evaluating the platform in 2026 are betting on an integration roadmap, not a finished product.

The financial structure of the deal — retiring $261.2 million in LivePerson debt alongside the $42.8 million equity price — gives the combined entity a clean balance sheet and removes a significant liability overhang. That matters for long-term stability, but it does not reduce the execution burden. SoundHound's claim to represent the "defining moment for the new agentic AI era," as CEO Mohajer put it, will be tested over the next 12 to 18 months as the company attempts to cross-sell voice AI into LivePerson's bank and airline relationships while delivering the seamless channel handoffs it is now promising.

Frequently Asked Questions

How much did SoundHound actually pay for LivePerson — and why is that different from the $43 million headline?

The $43 million figure represents only the equity consideration paid to LivePerson's stockholders. SoundHound simultaneously entered a notes restructuring agreement with LivePerson's secured creditors, paying approximately $261.2 million in cash and Class A common stock to retire LivePerson's outstanding debt at close. The total economic commitment was therefore approximately $304 million. Reporting the equity consideration alone without the concurrent debt retirement understates the deal's real cost by a factor of seven. The full financial terms are detailed in SoundHound's Q2 2026 10-Q.

What is OASYS and how does it differ from a standard enterprise chatbot?

OASYS (Orchestrated Agent System) is SoundHound's self-learning agentic AI platform. Unlike traditional chatbots or even first-generation AI agents — which require engineers to manually build, train, and maintain workflows — OASYS manages the full lifecycle of an AI agent fleet autonomously. It creates agents, orchestrates multiple specialized agents within a single customer interaction (dynamically selecting which sub-agent handles which step), evaluates their performance in production, identifies failure modes, and engineers its own updates for human expert review. The practical difference for an enterprise deploying it is that the system improves with real-world usage without constant engineering intervention. Full platform details were outlined in the OASYS launch announcement in May 2026.

Who are SoundHound's biggest competitors in enterprise AI customer service?

Salesforce (Agentforce Contact Center), Amazon Web Services (Amazon Connect + Lex), Genesys (which acquired Pinkfish in June 2026 for agentic workflow automation), Microsoft (Copilot Studio), and NICE (which acquired Cognigy for approximately $955 million in September 2025). The enterprise AI contact center market is consolidating rapidly, with multiple large-scale acquisitions in 2025 and 2026 as vendors race to capture footprint before the agentic AI market matures. Gartner projects enterprise agentic AI software spending will reach $985 billion by 2030.

Is SoundHound AI profitable, and when might the combined company break even?

SoundHound is not yet profitable. The company posted record Q2 2026 revenue of $61.9 million — up 45% year-over-year — but recorded a GAAP net loss of $42.8 million for the quarter and an adjusted EBITDA loss of $9.6 million. Company management has projected a path to profitable growth by early 2027 under the combined entity's cost discipline and cross-selling opportunity. Independent analysts tracked by MarketBeat carry a consensus Moderate Buy rating but have noted that many do not expect sustainable profitability until 2028.