Hear
Hear is an agentic conversation intelligence platform for contact centers, analyzing 100% of calls, chats, emails and tickets to automate quality assurance, flag compliance risk, surface revenue signals and supervise AI agents. Access is demo-only.
What is Hear?
Hear is an agentic conversation intelligence platform built specifically for contact centers. It is published by Intune Communication LTD, an Israeli limited company started in 2023 by Noam Fine (CEO) and Yossi Marouani (CTO), who had already co-founded four companies together, among them Sensiya, acquired by i.am+, and over.ai, acquired by Vonage.
The premise is straightforward. Most contact centers listen to somewhere between 1% and 5% of their interactions and steer the whole operation from that sample. Hear ingests all of them — inbound and outbound calls, web chat, email, WhatsApp and SMS, helpdesk tickets, CRM notes and digital support exchanges — then transcribes, diarizes and indexes each one. What comes out is presented as an interrogable record rather than a dashboard: a manager asks a question in plain language and gets an answer cited back to the exact moment of the conversation it came from.
Four stages structure the product — Ingest, Understand, Answer, Act — with four role-based modules on top, for customer experience, compliance, operations, and sales and marketing. Automated quality assurance scores every conversation against a customizable scorecard covering script adherence, objection handling, empathy and tone, upsell execution and first call resolution. A parameter extraction engine turns unstructured speech into reportable fields the customer defines, such as cancellation intent, price sensitivity or competitor mentions. Near real-time alerts surface compliance risk, churn signals and missed revenue.
A second product, AI Agent Monitoring, tackles a newer problem: supervising the AI agents a company has deployed. A dedicated monitoring model runs alongside every AI interaction and scores more than 60 behavioral signals across six dimensions — hallucination and grounding, security and adversarial behavior, quality and completion, compliance, customer experience, and efficiency.
Hear publishes SOC 2 Type II, ISO 27001:2022 and ISO 27799:2016 certifications together with a public trust center. It connects to Genesys Cloud, Freshdesk, Freshworks, Freshcaller, JustCall, Commbox, Pipedrive, Feefo, GreenStar and Trustpilot, and can also be purchased through Google Cloud Marketplace. There is no self-service sign-up and no published price: every route starts with a sales demonstration.
What it does
- Ingest and transcribe 100% of calls, chats, emails, messaging and tickets from your existing stack
- Score every conversation automatically against a quality scorecard you define
- Answer plain-language questions about the conversation record, with citations to the exact moment
- Detect compliance breaches, script deviations and missing disclosures across all interactions
- Flag churn risk and revenue opportunities from behavioral signals as conversations happen
- Monitor deployed AI agents on more than 60 behavioral signals without manual configuration
- Feed agent coaching with real, contextual examples instead of subjective feedback
When to use Hear / When not to
A quick filter to help you decide if Hear is the right fit.
When to use Hear
- Contact center and CX leaders running multi-site, multi-channel operations where manual call sampling covers only a fraction of interactions
- Quality assurance teams that want every conversation scored against a customizable scorecard instead of a 1-5% sample
- Compliance officers in regulated sectors — financial services, insurance, healthcare, telecommunications — who must evidence script adherence across all interactions
- Operations managers hunting the root causes of long handle times, repeat contacts and failed digital journeys
- Teams deploying AI agents in production and needing an independent layer that scores hallucination, adversarial behavior and policy violations
When not to use Hear
- Individuals and small teams: there is no self-service sign-up, no free plan and no published price — every route runs through a sales demonstration
- Buyers who need to compare costs before talking to anyone, since neither the website nor its sitemap carries a pricing page
- Developers looking to build on a documented public API: the architecture is described as API-first, but no documentation is published
- Anyone needing a mobile or offline product — Hear ships as a web console only, with no iOS, Android or browser extension
- Organizations that require an EU-established vendor with a designated GDPR Article 27 representative, which Hear does not name
How to use Hear
A typical end-to-end flow, from setup to results.
- Request a demonstration through the website, or contact the sales team directly — there is no self-service sign-up
- Alternatively, find Hear in the Google Cloud Marketplace catalog and subscribe with your existing GCP billing account
- Scope the deployment with the Hear team: channels to ingest, volumes, and which contact center systems are in play
- Connect the source platforms through native connectors or the API-based integration layer, ingesting transcripts, recordings, metadata and structured fields
- Define the business parameters to extract from conversations, such as cancellation intent, refund causes or competitor mentions
- Configure the quality assurance scorecards, weighting models and evaluation criteria that match your internal standards
- Set up alert rules and conversation classifications for compliance breaches, churn signals and revenue opportunities
- Set role-based access so each team sees only the conversations, reports and parameters it should
- Work day to day from the web console at console.hear.ai, querying the record in plain language rather than navigating dashboards
- Expect first actionable insights within weeks of initial ingestion, according to the FAQ, though the Google Cloud page promises days
Pros & Cons
Pros
- Full coverage of every interaction, against the 1-5% manual sampling that most quality programs settle for
- Three independently verified certifications — SOC 2 Type II, ISO 27001:2022, ISO 27799:2016 — and a public trust center detailing more than 40 controls
- Answers cited back to the exact moment of a conversation, so a claim can be checked rather than trusted
- A dedicated product for supervising deployed AI agents, a gap most conversation analytics platforms leave open
- Documented subprocessor lists, a Data Processing Addendum and the 2021 Standard Contractual Clauses incorporated
- Data stored on AWS European servers according to the privacy policy
- Purchasable through Google Cloud Marketplace, with unified billing and eligibility against committed use discounts
Cons
- No public pricing at all: no pricing page on the site or in its sitemap, so budgeting requires a sales conversation
- No self-service trial; the FAQ offers proof-of-concept engagements without ever saying they are free
- The terms and conditions page serves the privacy policy text verbatim, so no distinct contractual terms are actually published
- No postal address and no registration number published anywhere, despite the company declaring itself Israeli
- No stated position on whether customer conversations train AI models, and no documented opt-out
- No EU representative designated under GDPR Article 27, and the addendum still cites the invalidated EU/US Privacy Shield
- Headline performance figures are published without any methodology behind them
Pricing & Plans
Hear publishes no price. There is no pricing page on the website and none in its sitemap, no permanent free plan is mentioned, and every call to action routes to a demonstration request or to the sales team. The FAQ offers proof-of-concept engagements allowing a prospect to evaluate the platform against their own contact center data before committing to a full deployment, but does not state whether those engagements are free of charge. A second purchasing route exists through Google Cloud Marketplace, where the subscription is billed to the customer's existing Google Cloud account and counts toward their committed use discounts; the plan names and amounts are not disclosed on Hear's own pages.
Data, GDPR & hosting
A consolidated view of how Hear handles your data.
GDPR overview
GDPR references are concrete but not entirely consistent. The privacy policy concedes that Intune Communication LTD is not an EU-based organization and says its policy follows the spirit of the GDPR while complying strictly with Israeli privacy law. The FAQ goes further, describing processing under strict GDPR principles with the Irish Data Protection Commission named as supervisory authority. Tangible measures do exist: the 2021 Standard Contractual Clauses are incorporated (Module Two, Irish governing law), a Data Processing Addendum forms part of the terms and is published in the trust center, incidents are reported within 24 hours, and SOC 2 Type II plus ISO 27001:2022 and ISO 27799:2016 certifications are verifiable. Two gaps remain: no Article 27 EU representative is named despite the vendor being outside the EU, and the addendum still cites the EU/US Privacy Shield, invalidated in 2020.
Who owns the data?
Hear's privacy policy states that personal data remains the property of the disclosing party, and its embedded Data Processing Addendum casts the roles explicitly: the customer is the data controller, Hear.ai acts as processor. The customer therefore stays responsible for lawfulness, consent records and individual rights requests, while Hear commits to strict security and confidentiality obligations and to reporting incidents within 24 hours. The publisher, Intune Communication LTD, states that it does not sell personal data. Access is limited to employees and contractors under monitored confidentiality obligations, and to named third-party providers bound by equivalent commitments. The documents carry an effective date of 1 January 2025.
Reuse rights
Hear says it collects contact details, payment details and contact-form content, plus non-identifying activity and technical data, and that it uses them to deliver the service and for what the policy calls continuous improvement. It undertakes not to sell that data and not to disclose it for direct marketing. Sharing is limited to named subprocessors: AI models (Microsoft Azure, Anthropic, OpenAI), analytics (Plausible Analytics, Hotjar, Pipedrive, Mixpanel, Google Analytics, Datadog), hosting (Landen, AWS), messaging (Reply.io, MailChimp, SendGrid) and payments (Stripe). Customers can access, correct, delete and port their data by email or through the contact form, and can withdraw marketing consent at any time. What the policy never states is whether conversation data feeds model training: the phrase continuous improvement sits next to three AI model providers without a stance being taken, and no opt-out mechanism is documented anywhere on the site.
Data retention & training
Hosting summary
The privacy policy is explicit on one point and vague on the rest: data is stored on AWS European servers, and some cookies are hosted outside the EU. No host country is ever named, only the region. The trust center published through Scytale gives a partly different picture, listing GCP as the cloud provider and MongoDB as the database system, with no region attached. Hear can also be deployed inside a customer's own Google Cloud environment when purchased through Google Cloud Marketplace, and the FAQ claims compatibility with cloud, hybrid and on-premises architectures, which means the effective location may depend on the deployment model agreed with the vendor. On jurisdiction, the publisher is an Israeli company; the 2021 Standard Contractual Clauses are incorporated under Irish governing law, and the FAQ names the Irish Data Protection Commission as supervisory authority. A prospective buyer should ask for the hosting location to be stated contractually rather than inferred from these pages.
Things to keep in mind
Risks and trade-offs to weigh before adopting Hear.
- The terms and conditions page serves the privacy policy word for word, so the contractual terms you would be agreeing to are not actually published — ask for them in writing before signing
- The vendor is established outside the EU and designates no GDPR Article 27 representative, while its addendum still cites the EU/US Privacy Shield, a framework invalidated in 2020
- Nothing on the site states whether your conversations train AI models, and no opt-out is documented, even though three AI model providers are named as subprocessors
- Two different subprocessor lists coexist and barely overlap: the privacy policy names AWS and Landen for hosting, the trust center names GCP and MongoDB — check which one governs your data
- Recording, transcribing and scoring 100% of interactions has consequences for your own agents: the vendor frames it as coaching rather than surveillance, but that boundary is set by how you configure it, not by the tool
- Headline figures such as 142 million calls analyzed, 340+ enterprise deployments or 11.4 million dollars of recovered revenue are published without any methodology, and should be treated as vendor claims
- Setup timelines contradict each other across the site — hours, days or weeks depending on the page — so plan against the most conservative one
Setup & Integrations
Technical difficulty
Moderate, and never a solo effort. There is no self-service onboarding: deployment begins with a sales conversation and is then run with the Hear team, covering integration setup, custom parameter definitions, quality scorecards, dashboards and reporting logic. Connection happens through native connectors or the API-based integration layer, ingesting transcripts, recordings and metadata without changing existing workflows, and SSO is mentioned. The technical burden therefore sits mostly with the vendor, but the configuration work — deciding what to extract and how to score it — is real and belongs to the customer.
Deployment
Integrations
Behind Hear
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What exactly does Hear do?
How much does Hear cost?
Is there a free trial?
What security and compliance certifications does Hear hold?
Where is the data hosted?
Are customer conversations used to train AI models?
Which systems does Hear integrate with?
Who is behind Hear, and how is it funded?
Should you pick Hear?
Hear makes a clear, defensible promise: stop steering a contact center from a 1-5% sample and work instead from every conversation your customers actually have. The execution matches the promise on the points that can be checked. Three independent certifications, a public trust center detailing more than 40 controls, a named subprocessor list, an embedded Data Processing Addendum and the 2021 Standard Contractual Clauses put it well above the average of tools in this category. The natural-language querying with citations back to the exact moment of a call is a genuine design decision rather than a slogan, and AI Agent Monitoring addresses a supervision gap that most conversation analytics vendors have not yet touched.
The reservations are just as concrete. Nothing about the commercial side is public: no price, no plan names, no self-service path, and a proof-of-concept offer whose cost is never stated. The legal layer has loose ends that matter for a European buyer — the terms and conditions page serves the privacy policy verbatim, so no contractual terms are actually published; no Article 27 EU representative is named even though the publisher openly declares itself outside the EU; the addendum still cites the EU/US Privacy Shield, invalidated in 2020; and the site never says whether customer conversations feed model training, while naming three AI model providers as subprocessors.
This is enterprise software for organizations that already run a serious contact center and can put a procurement and security review behind the purchase. Those buyers will find plenty of substance and a vendor willing to document itself. Everyone else should know the door only opens through a sales conversation, and that several answers a careful reviewer needs are not on the website today.
- Choosing a selection results in a full page refresh.
- Opens in a new window.