Verint
Also known as: Verint Systems, Verint Intelligent Virtual Assistant, Verint IVA, Verint Conversational AI Agents, Verint Agent Factory, Calabrio
Verint's Conversational AI Agents resolve customer requests across voice and digital channels, including payments, reservations and account changes. Agent Factory coordinates specialized agents, shared context and human handoffs, with configurable models and industry assistants as starting points. Knowledge Management supplies maintained service content, while QM Intelligence evaluates automated conversations. The software can sit alongside an existing contact center stack. Verint is owned by Thoma Bravo and now includes Calabrio; agent packages and implementation are sold through sales.
Verint Systems, Inc. sells customer experience automation software, and its Conversational AI Agents, coordinated by Agent Factory, are its autonomous customer service layer. Thoma Bravo completed its acquisition of Verint on November 26, 2025 and combined it with its portfolio company Calabrio; the combined business runs under the Verint brand, and Calabrio products carry on with no forced migration. Workforce management, recording, quality and analytics products sit around the agents on the same platform.
Conversational AI Agents handle voice, chat and messaging through the customer's existing contact center systems and complete tasks end to end: authenticating a customer, processing a payment, updating an account or changing a reservation, and escalating with full context when a person is needed.
Agent Factory orchestrates prebuilt and custom agents, passing interaction history between them and routing to staff where judgment matters, with centralized prompt management, model governance and support for MCP, A2A and ACP. Customers can use hundreds of models from OpenAI, Google Gemini, Anthropic and Meta Llama or bring their own.
Industry agents for healthcare, travel, financial services, retail and the public sector deploy from IVA Studio, a no code tool for designing, testing and updating workflows, and Knowledge Management supplies a maintained, search optimized repository shared with human agents.
Pricing and implementation come through a sales proposal, with no public entry price, unit rate or included allowance; support plans, professional services and starter packages are scoped separately, and the platform targets 99.99% availability.
Users authenticate through SAML 2.0 or OpenID Connect with role based access, data is encrypted at rest and in transit, and the certification matrix lists the voice and digital virtual assistant under SOC 2 Type II coverage. The platform runs in Verint's public cloud with data kept in the hosting region, in a customer's private cloud, or in a hybrid that keeps recordings at the customer's site. A library of platform APIs for provisioning, interaction data, workforce schedules, KPIs and transcription sits behind a customer developer portal.
Verint fits large contact centers that want automated resolution, human escalation and quality oversight in one platform, especially those already running Verint or Calabrio workforce products. QM Intelligence scores AI agent conversations on the same forms as human ones and explains each score, which gives quality teams a direct handle on automated service. The gaps are agent memory across contacts, approval of individual actions before they run, a step by step record of tool calls and a public API for the agent layer, and commercial terms depend entirely on the agreed scope.
Vendor details
Canonical URL
https://www.verint.com
Category
Customer support agent
Subcategory
Conversational AI Agents coordinated through Agent Factory
Funding status
Privately held under Thoma Bravo following completion on November 26, 2025. Calabrio joined the combined Verint business, operating under the Verint corporate brand in 2026.
Company status
independent
Use cases & customers
Primary use cases
Target customers
Deployment options
Integrations
Agent Factory connects agents to enterprise data and tools, including MCP compatible integration paths. Channel Automation supports existing telephony and digital messaging. Platform APIs cover identity provisioning, interaction enrichment, schedules, KPIs and transcription through a customer developer portal. Knowledge Management feeds persistent service content into IVA workflows.
In practice
A traveler requests a reservation change; the agent checks account context, completes the change in connected systems and transfers an exception with the conversation history.
A financial services customer enters a payment workflow that verifies identity, processes the transaction and confirms completion.
A quality team scores automated conversations against its evaluation forms, inspects the source moments behind failed answers and revises the service workflow.
Sources & related URLs
Related / legacy domains
Research sources
Agentic Index coverage score
11.5 / 14 capabilities · 82%
| Integrations & Tool Calling | Full |
|---|---|
|
Conversational AI agents complete transactions in connected systems: authenticating a customer, processing a card payment, updating an account or changing a reservation, then confirming the result. Agent Factory links prebuilt and custom agents to enterprise data and tools, with support for MCP, A2A and ACP and governance controls over what each agent can reach, and the open ecosystem connects business applications and channels on top of the customer's existing stack. Which connectors a deployment uses, and the credentials behind them, are set up per customer during implementation. Sourceverint.com/agent-factoryread 2026-10-19 |
|
| Workflow Orchestration | Full |
|
Agent Factory orchestrates a workforce of prebuilt and custom AI agents and people, deciding which agent handles each step of a request, such as authentication, account lookup and resolution, and passing the interaction history along at every handoff. The same prompt management, model governance and policy guardrails apply across every agent in a workflow, and work routes to a person wherever judgment is needed. Voice and digital agents take part in the same flows, alongside the virtual agents and assist tools a contact center already runs. Sourceverint.com/agent-factoryread 2026-10-19 |
|
| Knowledge Grounding & RAG | Full |
|
Knowledge Management keeps a shared repository of policies, procedures and service content that feeds both virtual assistants and human agents. New material is analyzed and optimized for search as it arrives, and natural language retrieval uses customer context to pick the relevant answer. Knowledge Automation can also search content where it already lives in other enterprise systems, so agents can use it before, or without, a migration into the repository. Sourceverint.com/knowledge-managementread 2026-10-19 |
|
| Human Oversight & Guardrails | Partial |
|
Workflows hand an interaction to a person when judgment, risk or complexity calls for it, passing the conversation context, and Agent Assist copilots can support the employee who takes over. Central prompt governance and policy guardrails limit what agents do on their own. A payment or record change the agent proposes is not held in a queue for a reviewer to approve before it runs; oversight of finished conversations comes afterward through quality scoring. Sourceverint.com/agent-factoryread 2026-10-19 |
|
| Security, Identity & Governance | Full |
|
Users sign in through SAML 2.0 with a customer supplied identity provider or through OpenID Connect, role based access with least privilege governs the cloud platform, and API calls are checked against Verint Identity Service. Recordings and databases use AES 256 encryption, and TLS protects data in transit. The product certification matrix lists voice and digital Intelligent Virtual Assistant under annually audited coverage that includes SOC 2 Type II; its HITRUST entry maps the SOC report and is not a HITRUST certification, and detailed reports sit in the trust center. Sourceverint.com/wp-content/uploads/2026/03/verint-certification-product-matrix.pdfread 2026-10-19 |
|
| Observability & Auditability | Partial |
|
Dashboards track service performance across human and automated conversations, and quality scores come with the AI reasoning tied to the moment in the interaction behind each score, so staff can find recurring problems. What an agent did inside a workflow, which tool it called, with what arguments and what came back, has no step by step record a team can open, so a failed transaction is traced through conversation review and the connected systems' own logs. Sourceverint.com/quality-and-complianceread 2026-10-19 |
|
| Memory & State Persistence | Partial |
|
Interaction history and the CX Data Hub carry customer context between specialized agents and the staff who take over, and a channel switch keeps the conversation intact. Historical engagement data also feeds training and analysis. Agent Factory has no memory layer of its own with a stated scope, lifetime and controls for facts an agent learns, so continuity from one contact to the next comes from the shared data foundation and saved conversations. Sourceverint.com/agent-factoryread 2026-10-19 |
|
| Deployment & Data Residency | Full |
|
The Verint platform runs in Verint's or a partner's public cloud, in a private cloud belonging to the customer or its partner, or in a hybrid of the two. In Verint's cloud, customer data and backups stay in the region where the service is deployed, spread across at least two isolated zones on one of the three major cloud providers, and the Hybrid Cloud model keeps audio and screen recordings, and optionally transcription, at the customer's telephony site. These options are set out for the platform as a whole, without saying which Agent Factory and model components run in a customer cloud. Sourceverint.com/technical-architecture-overviewread 2026-10-19 |
|
| Prebuilt Agents / Templates / Packs | Full |
|
Industry agents for healthcare, travel, financial services, retail and the public sector deploy with one click from IVA Studio, a no code tool for designing, testing and updating workflows, and are then adapted to local systems and policies. Their tasks range from appointment booking and prescription refills to reservation changes, bill payments and returns. Agent Factory also offers prebuilt agents for common customer service use cases, which combine with a customer's own custom agents in one governed workflow. Sourceverint.com/industry-specific-ai-agentsread 2026-10-19 |
|
| Triggers & Channel Coverage | Full |
|
Voice calls and digital messages reach conversational agents through the customer's existing IVR, contact center and messaging connections, with no rip and replace. Channel Automation routes each interaction on intent, workload, schedules and earlier virtual assistant context, and customers can move between web chat, messaging and telephony within one service journey. Calabrio customers can switch on Verint's bots on their current setup without a migration or new infrastructure. Sourceverint.com/channel-automationread 2026-10-19 |
|
| Model Flexibility & Routing | Full |
|
Agent Factory gives customers access to hundreds of models from providers such as OpenAI, Google Gemini, Anthropic and Meta Llama, or a model of their own, configured and managed in one environment alongside centralized prompts and model governance. Separately, Da Vinci picks models internally for packaged Verint services, so the choice applies to agents built and run in Agent Factory and may not reach every prebuilt bot. Sourceverint.com/agent-factoryread 2026-10-19 |
|
| APIs / SDKs / MCP Extensibility | Full |
|
The API Gateway exposes cloud platform APIs with throttling and access enforced through Verint Identity Service, covering user provisioning, live and post call interaction enrichment and export, workforce schedules and adherence, performance KPIs and the Da Vinci Speech Transcription API. The full reference sits in the Verint Developer Portal behind a customer login. These operations cover the wider platform; building or running Agent Factory agents is not among them, and the MCP, A2A and ACP support describes how agents connect outward to other systems. Sourceverint.com/technical-architecture-overviewread 2026-10-19 |
|
| Testing, Debugging & Optimization | Full |
|
QM Intelligence scores up to 100% of voice and digital interactions, for AI agents as well as people, against the customer's own evaluation forms or a question library, and Automated Answer Validation checks that answers were correct and compliant and recorded accurately in the CRM. Each score shows the AI reasoning tied to the moment it came from, so teams can see why an automated conversation failed and revise the workflow, and IVA Studio lets builders test a workflow before deploying it. Sourceverint.com/quality-and-complianceread 2026-10-19 |
|
| Browser / Computer-use | Not documented |
|
Desktop Intelligence captures what employees do in their applications and uses a Process Discovery Bot to map workflows and recommend improvements, but people still operate those screens. Conversational AI agents complete tasks through connected systems and workflow orchestration, with no browser or desktop session of their own to click, type and navigate through another application when an integration is missing. Sourceverint.com/desktop-intelligenceread 2026-10-19 |
|
The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded
Pricing
Contact sales
License, usage and service terms defined in the customer proposal
Included quota
No public included conversation, minute, model or agent allowance for the current agent package.
What is public
A sales route for the agent products, modular deployment approach and implementation service options.
Billing mechanics
The proposal defines licensing, usage, included capacity and any additional service charges.
Cost watchouts
Implementation, integrations, telephony connectivity, additional modules and ongoing support can raise the total, and whether each is included in the agent package depends on the proposal.
Variable cost rationale
Public offers do not specify the billing unit, model usage treatment, included allowance or overage rules.
Additional watchouts
The public offer has no unit rate or bundled allowance, and existing contact center systems retain their own costs.
Overage / add-ons
No public overage tariff or automatic recharge rule for Conversational AI Agents or Agent Factory.
Sales call required
Yes, required for paid access
Free / trial
Sales demo available; no public free tier or self service agent trial allowance.
Lowest paid plan
Custom sales proposal for Conversational AI Agents and Agent Factory.
Commercial notes
Agent Factory and Conversational AI Agents can be introduced alongside existing contact center systems. Scope and implementation are agreed with sales.
Key ambiguities
No public entry price, unit rate, allowance, minimum commitment or Agent Factory license structure; older analyst reprints of IVA pricing describe a previous offer.
Cancellation / refund
Contract length, renewal, cancellation and refund provisions depend on the customer agreement; website terms do not establish commercial subscription terms.
Support SLA / resale
The architecture offer includes support plans and separately scoped professional services. The platform availability target is 99.99%, excluding planned maintenance; contractual credits and agent package coverage require the order terms.
Missing data
Entry fee, usage units, quotas, overages, setup price, minimum term, volume discounts, cancellation terms and trial allowance.
Related vendors
- Decagon — AI agent platform purpose-built for customer support, with deep…
- 5.Y — 5
- Ada — Agentic customer experience platform whose AI agents autonomously…
- Aisera — Enterprise agentic AI platform that orchestrates specialized agents…
- ASAPP — Generative AI for enterprise contact centers that plugs into…
- Assembled — AI customer support orchestration platform that unifies autonomous…