Chattermill

Chattermill is a London-based customer experience intelligence (CXI) software company that unifies customer feedback from many channels — surveys, online reviews, contact-centre and support conversations, social media and product feedback — and applies AI to tag, categorise and cluster it into insights. Signals are enriched with business context such as customer ID, channel and location, and layered with CRM, operational, transactional and competitive data so that findings can be tied to metrics the business already tracks, such as churn, retention and spend. Outputs include precision insights on specific issues, impact/driver analysis, custom reports and dashboards, anomaly detection and real-time alerts routed to Slack, email, Jira and similar tools.

The company markets a proprietary model, Lyra, described as trained specifically for customer experience intelligence, and a 'CXI agentic architecture' composed of a data layer, a CX context engine combining frontier LLMs with fine-tuned specialist models, an analytical engine, workflow and alerting, insight governance, and a skills-and-evaluations library. An MCP interface lets external agents such as Claude and ChatGPT query Chattermill data and generate downstream artefacts (for example Linear tickets or Notion documents). Several components — Lyra Agent, knowledge and memory, collaboration and audit logging, and a CX Context Graph — are listed on the website as not yet released.

At the time of its 2022 Series B, Chattermill described itself as a 'Unified Customer Intelligence' platform using deep-learning AI, ingesting data both from internal sources such as support tickets and from public web sources, and integrating with survey tools including SurveyMonkey, Typeform and Qualtrics. Press coverage states the platform analyses feedback natively in 50+ languages, and CEO Mikhail Dubov has framed the company's goal as analysing over a billion pieces of customer feedback for clients.

Founding story

Co-founders Mikhail Dubov and Dmitry Isupov met through Entrepreneur First and founded Chattermill in London in 2015. Isupov had worked at a market research company where customer feedback and brand perception were reviewed manually; the founders concluded that such data held significant insight but was expensive, difficult and slow to process because it was human-driven, with research cycles taking weeks or months. They set out to achieve the same outcome with machine learning, betting that automated reading of all feedback — rather than sampling — was the hard and valuable part of the problem.

Business model

B2B SaaS sold to enterprises, with a demo-led, sales-qualified entry point (no public pricing; prospects are asked about monthly feedback volume, banded as under 1,000, 1,001–5,000, or more than 5,000 items per month). The platform is positioned as enterprise-grade, citing SOC 2 Type II, ISO 27001 certification, GDPR/CCPA compliance, SSO, access controls and PII redaction. Alongside software, Chattermill offers a dedicated expert team under an 'XLG Acceleration' service that links customer signals to retention, churn and revenue outcomes.

Software subscription sold to enterprises via a book-a-demo, quote-based motion; pricing is not published and prospects are segmented by monthly feedback volume. Additional value is delivered through an expert services offering (XLG Acceleration) tying CX signals to business outcomes.

Traction

Named enterprise customers include Uber, Amazon, H&M, HelloFresh, Tesco, Zappos, Just Eat, Qonto and E.ON Next. At Series B the company reported hitting 200% of new business revenue and pipeline targets in Q1 2022 and said customers such as Uber and Just Eat had tripled usage. Public review platforms show 4.5/5 ratings on G2 (236 reviews) and Gartner (92 reviews) as of 2026, with the site citing over 200 five-star reviews. Customer-cited outcomes on the company's site include a 42.8% reduction in contacts per transaction and a 144% NPS increase (from +9 to +22), and one customer reference notes eight years working with Chattermill.

Latest developments

As of 2026 the website presents Chattermill as an 'AI-native' CX intelligence platform organised around four pillars (Data, Intelligence, Agent, Program), with the Lyra proprietary model, an MCP-and-skills integration for third-party AI agents, and a CXI agentic architecture. Several architecture components are labelled coming soon, including a Chattermill-native Lyra Agent, a knowledge-and-memory layer, collaboration and audit logging, and a CX Context Graph. The site lists a broader executive team beyond the founders, including a Chief Revenue Officer, Chief Scientist and VPs across engineering, product, finance, people, marketing and XLG, and cites 2026 analyst-community recognitions from G2 and Gartner.