SINGAPORE, May 7, 2026 /EINPresswire.com/ — Revelir AI, a customer service AI company, today announced the launch of Revelir Insights, an AI analyst designed to explain why customer experience (CX) metrics move. The new product connects directly to Anthropic’s Claude, allowing CX leaders to query 100 percent of their support dataset in plain English.
Every month, support teams at fintechs, travel platforms, and e-commerce companies handle tens of thousands of customer conversations. While manual quality assurance (QA) programs typically review only 1 to 5 percent of these interactions, the remaining data contains critical signals about product friction and churn risk. Revelir Insights solves this by reading every support conversation, enriching each one with structured AI signals, and connecting the resulting dataset to Claude through the Model Context Protocol (MCP).
“Most CX leaders can tell you their CSAT score. Very few can tell you why it moved,” said Rasmus Chow, Founder of Revelir AI. “The data to answer that question exists in every support conversation, but until now nobody could query it at scale. Revelir Insights changes that. The AI analyst layer is the difference between knowing a metric moved and knowing why.”
Core Capabilities of Revelir Insights
The platform introduces four primary capabilities for support operations:
1. Plain-English Queries Across the Full Dataset: Through the MCP integration, a Head of Customer Experience can ask Claude questions such as, “What drove the spike in negative sentiment last week?” Revelir Insights answers in seconds, citing the underlying tickets as evidence.
2. AI-Enriched Signals on Every Ticket: The system evaluates every conversation for sentiment arc (how the customer felt at the start versus the end), contact reason classification, and product feedback. Even tickets marked “resolved” by the helpdesk are analyzed for retention risk.
3. Custom Metrics Defined in Plain English: CX and product leaders can define their own AI-inferred metrics, such as compliance checks or customer effort indicators. These custom signals become live, queryable fields available to Claude through the MCP connection.
4. Live Signal Feed for Churn Risk and Feedback: The platform surfaces a continuous feed of churn-risk signals and sentiment drivers extracted from incoming tickets, turning support data into a real-time signal for product and operations leaders.
Enterprise Production Foundation
Revelir Insights sits on top of the same enrichment layer the company has been running in enterprise production for over a year through its companion product, RevelirQA. This automated QA engine is already deployed at major Southeast Asian enterprises.
At Xendit, one of Southeast Asia’s leading payments infrastructure providers, RevelirQA evaluates every support ticket against a compliance-led scorecard, providing a full audit trail per score. This has allowed the QA team to shift from manual sample-grading to AI calibration and pattern-based coaching across the full ticket population.
At Tiket.com, one of Indonesia’s leading online travel platforms, RevelirQA operates across a high-volume multilingual environment in Bahasa Indonesia and English, covering both chat and voice channels. The system scores 100 percent of tickets without requiring additional QA headcount.
Both deployments process thousands of conversations per week, generating the structured data foundation that Revelir Insights now queries.
About Revelir AI
Revelir AI is a Singapore-based customer service AI company. It builds an automated QA engine that scores 100 percent of support conversations, and an AI analyst (Revelir Insights) that turns the resulting data into answers about why CX metrics move. Enterprise customers include Xendit and Tiket.com. The platform integrates primarily with Zendesk, with API access available on Professional and Enterprise plans.
Rasmus Chow
Revelir AI
email us here
Legal Disclaimer:
EIN Presswire provides this news content “as is” without warranty of any kind. We do not accept any responsibility or liability
for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this
article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
Media gallery
