We believe

    Agents don't need a better model.They need a better AI data stack.

    An AI agent can't work well on fragmented, untrusted data. We are a boutique data consultancy that designs, builds and modernises governed data stacks for reliable reporting and agentic analytics.

    45 minutes · no obligation · you'll speak with a founder, not sales

    Trusted by data teams at

    • Hunkemöller
    • Carwow
    • Pleo
    • ROUVY
    • Barton Peveril
    • BARK
    • Osprey
    • RIXO
    • Cyanocapture
    • Ageras Group
    • Breakthrough
    • BridgeU
    • ColourPop
    • Conversion Interactive
    • Florence
    • FSG Smart Buildings
    • Insighta
    • INTO University Partnerships
    • Popsa
    • Qubit
    • Rebtel
    • Torticity
    • Welbee

    >50%

    of business decisions will be driven by AI agents by 2027, according to Gartner.

    Those agents will need data they can rely on, and most businesses don't have it in one place yet.

    Why it matters

    Why do agents need a better AI data stack?

    1. So agents give the same answer as your dashboards, and people can trust it.
    2. So agents can work across your whole business, not one system at a time.
    3. So every answer can be traced, tested and governed before anyone acts on it.
    A Hunkemöller store colleague using Looker on a tablet with a customer

    A better AI data stack

    What would a better AI data stack give you?

    • The right answer
    • In the right context
    • for the right person or agent

    The right answer, in the right context, for the right person or agent.

    What we build

    A better, context-aware, AI data stack.

    Every answer, to a person or an agent, comes from the same governed source.

    Your sources

    • Product events
    • CRM and marketing
    • Finance and ops
    • SaaS apps

    Five layers, built and governed end to end

    • ServeDashboards for people. MCP tools and APIs for agents.
    • GovernTests, lineage, access controls and evals on every answer.
    • DefineOne semantic layer. Every metric defined once, in business terms.
    • ModelClean, documented models of customers, products and revenue.
    • IntegrateEvery source system loaded, monitored and reconciled.

    Same answer for

    People

    Trusted dashboards and self-service

    Agents

    Conversational analytics over MCP

    Which channel drove the most revenue last month?

    Paid search: £148k, 36% of net revenue.

    The path to trusted data

    What does the path to a better AI data stack look like?

    Each step builds on the one before. Here is what each one has been worth to our clients.

    Step 1 · Stop arguing

    One set of numbers

    Revenue, customers and margin defined once and agreed by every team.


    Pleo

    60%

    cut in time to insight, from one set of numbers the whole team trusts.

    Step 2 · Stop waiting

    Answers on demand

    Teams answer their own questions from dashboards they trust. Analysts move on to the hard problems.


    UK College

    8 days

    from signed contract to live dashboards and trained users, for £7.1k.

    Step 3 · Get ahead

    Put AI to work

    AI assistants and agents use your trusted data to answer questions and take action. Forecast who will buy, stay or leave.


    US Agency

    New revenue

    The client now sells analytics to its own customers, built on the platform we delivered.

    Who we are

    Introducing Rittman Analytics

    An independent analytics consultancy founded in 2018 by Mark Rittman and Lewis Baker. Based in the UK, working with data-rich businesses across the UK, EU and USA.

    Who we are

    AI-first engineers, part of your team

    • Mark Rittman
    • Lewis Baker
    • Tim Griew
    • Alex Caldwell
    • Jordan Ilyat
    • Olivier
    • Cesar
    • A team of certified Google Cloud consultants
    • Specialists in Looker and the BigQuery stack
    • 10+ years delivering modern analytics

    Our partner network

    A stack we know inside out

    Rittman Analytics partner logos including Google Cloud, dbt Labs, Cube, Fivetran, Anthropic, Snowflake and Dagster

    Our delivery methodology

    Wire: AI-accelerated delivery

    Wire
    • AI generates the boilerplate: models, tests, docs and dashboard prototypes
    • Our consultants keep the architecture, governance and judgment

    45 minutes · no obligation · you'll speak with a founder, not sales

    Our methodology

    It's never just a technical problem.

    Lasting change depends on the team and ways of working around the platform, not just the platform. Every engagement follows the same three steps.

    1. 01 · Confront

      Confront the hard truths

      A diagnostic-first look at your data, tooling and ways of working: a clear picture of what's holding you back.

    2. 02 · Create

      Create the change

      We build a modern, sustainable data foundation: the warehouse, the models and the semantic layer your team can rely on.

    3. 03 · Manage

      Manage the change

      Embedded enablement and ongoing support, so the transformation sticks long after our engagement ends.

    Case study: Hunkemöller built an AI-ready data platform and team with Rittman AnalyticsCase study: Pleo and Rittman AnalyticsCase study: ROUVY and Rittman AnalyticsCase study: Carwow and Rittman AnalyticsCase study: Cyanocapture and Rittman AnalyticsCase study: Barton Peveril and Rittman AnalyticsCase study: Osprey and Rittman AnalyticsCase study: RIXO and Rittman AnalyticsCase study: BARK and Rittman Analytics
    Proof