The short version
Salesforce is buying Cresta to make voice a native channel for its AI agents. The deal, announced on November 2, 2026, adds AI voice agents, real-time coaching and conversation analytics to Service Cloud, with closing expected in February–April 2027. Teams that run support on Salesforce should plan their Salesforce integration and telephony roadmap with this in mind.
What exactly did Salesforce announce?
Salesforce said it has signed a definitive agreement to acquire Cresta, which it describes as a voice-first operating system for customer experience. Cresta’s platform runs voice agents that handle complex calls from start to finish, guides human representatives with suggestions during live conversations and analyzes interactions after the fact. Salesforce says Cresta already runs at enterprise scale for brands including Marriott, Comcast and Alaska Airlines.
After closing, Cresta will be integrated with Service Cloud so that every call is tied to the customer record and to the outcome it drives, which Salesforce frames as resolution, retention and revenue. “Voice is where service is hardest to get right, and where it matters most to customers,” said Kishan Chetan, executive vice president and general manager of Agentforce and Service Cloud. Salesforce did not disclose financial terms. Reuters, citing The Information, reported the price at about $2.7 billion.
Why is Salesforce buying a voice AI company now?
Cresta is the latest step in an unusually busy year of customer experience deals. In June, Salesforce agreed to buy AI customer service platform Fin for $3.6 billion, a deal CNBC described as a push to strengthen its agentic offerings. CMSWire counts Contentful, Fin, Listen Labs, Braze and now Cresta among Salesforce’s 2026 acquisitions. Fin brought a chat-first AI agent; Cresta fills the voice gap, where latency, accents, interruptions and compliance scripts make automation much harder than in text.
Cresta started as a project at Stanford’s AI Lab and was co-founded in 2017 by Sebastian Thrun, among others. Its pitch combines full automation with support for human agents, which matters because most large contact centers will run hybrid models for years: AI takes routine calls, people take the hard ones, and both need the same context. COMMfusion analyst Blair Pleasant summed up the fit to CMSWire: “Cresta understands the conversation, Salesforce understands the customer.”
What it means for US & EU software teams
For companies building on Salesforce, the practical effect is that voice AI moves from a bolt-on vendor to a first-party option. That simplifies procurement, but it also deepens lock-in: the agent logic, call transcripts and outcome data will increasingly live inside one platform. Teams that design AI agents for service should keep business rules, knowledge sources and escalation logic in systems they control, and connect them to the CRM through documented APIs rather than platform-only configuration.
Voice also raises the compliance bar. Call recordings and transcripts carry personal data, and in the EU automated handling of customer calls brings GDPR duties on transparency, retention and data transfers, plus transparency obligations for AI systems that interact with people under the EU AI Act. In the US, consent rules for recording vary by state, and FinTech and HealthTech teams add PCI DSS and HIPAA requirements on top. Whoever builds the integration needs to decide early where audio is processed, how long it is stored and how a caller reaches a human.
What should teams do before the deal closes?
- Map your voice stack. List which telephony, contact center and analytics tools touch calls today and how they connect to customer records in Salesforce.
- Keep data portable. Store transcripts, intents and call outcomes in formats and systems you can export, not only inside a vendor dashboard.
- Separate logic from platform. Put routing rules, knowledge bases and escalation policies behind your own APIs so voice and chat agents share them.
- Review compliance early. Confirm recording consent, retention periods, data residency and AI disclosure for each market before scaling automated calls.
- Pilot on a narrow use case. Start with one high-volume, low-risk call type and measure resolution, handle time and transfer rate against human-only handling.
Frequently asked questions
What did Salesforce announce about Cresta?
On November 2, 2026, Salesforce said it signed a definitive agreement to acquire Cresta, a voice AI platform for customer experience. Cresta runs voice agents that resolve calls end to end, guides human representatives in real time and analyzes customer interactions. After closing it will be integrated with Service Cloud.
How much is Salesforce paying for Cresta?
Salesforce did not disclose the price. The Information reported that Salesforce agreed to pay about $2.7 billion, and Reuters carried that report. Until official figures appear in filings, treat the number as reported rather than confirmed.
When will the acquisition close?
Salesforce expects the deal to close in the first quarter of its fiscal year 2028, which runs from February to April 2027. Closing is subject to regulatory approvals and customary conditions. Cresta will keep supporting the contact center and telephony platforms its customers already use.
What does this mean for companies that use Salesforce for customer service?
Voice AI agents are likely to become a first-party Service Cloud option, tied directly to customer records. That can shorten projects, but it also concentrates call data and agent logic in one platform. Keep business rules, knowledge sources and call data portable, and review GDPR, recording consent and AI disclosure rules before scaling automated calls.
Sources
Salesforce — Salesforce Signs Definitive Agreement to Acquire Cresta (primary source, November 2, 2026)
Reuters — Salesforce acquires startup Cresta for $2.7 billion, The Information reports (November 2, 2026)
CMSWire — Salesforce adds Cresta to the cart in its 2026 CX shopping spree (November 2026)
CNBC — Salesforce to buy AI customer service platform Fin for $3.6 billion (June 15, 2026)