Skip to content

DataGardener MCP

UK Company Intelligence for AI Agents

From Artificial Intelligence to Actual Intelligence.

DataGardener MCP connects AI agents to trusted UK company, financial, lending, procurement, people, risk, property and economic intelligence through one secure endpoint. Every answer carries its source and date.

Explore the MCPBook an MCP demo
View developer overview
UK company records
17M+
verified data sources
40+
Refresh
Daily

Choose a question

Question

Find engineering SMEs in the West Midlands that are growing, have won public-sector work and may need additional working capital.

Stage 1 of 5

Understanding the request

The agent breaks the question into the conditions it must satisfy.

  • Geography: West Midlands
  • Sector: engineering
  • Size: SME
  • Growth signal
  • Public contracts
  • Working-capital signal

Each condition maps to one or more DataGardener intelligence layers. Nothing is guessed: if a condition cannot be evidenced, the agent says so.

  • Company Intelligence
  • Financial Intelligence
  • Procurement
  • Lending
  • Risk
Read this investigation as text

Question: Find engineering SMEs in the West Midlands that are growing, have won public-sector work and may need additional working capital.

Intent: Geography: West Midlands; Sector: engineering; Size: SME; Growth signal; Public contracts; Working-capital signal.

Tools selected: search_companies, execute_ukgaap_query, search_contracts, execute_query.

Modules queried: Company index (Engineering SIC codes, West Midlands, live, SME size bands: 1,284 candidates); UK GAAP financials (Turnover up over two periods, cash down or trade debtors rising: 216 match); Contract Finder index (Awarded supplier in last 24 months, any buyer: 41 match); Company index (No new charge in 36 months, no insolvency event: 27 shortlisted).

Evidence connected: Turnover growth and debtor build-up from two consecutive filings; Contract award notice with buyer and value; Charge history showing age of last secured facility; Company status, incorporation date and SIC description.

Answer: 27 engineering SMEs in the West Midlands match all four conditions. The strongest signal set combines two years of turnover growth with rising trade debtors, a public-sector award in the last 24 months, and a last registered charge older than three years.

Provenance: Source Companies House filings; Contracts Finder; date Filed accounts to 2025; awards to 2026; module Company, Financial, Procurement, Lending; usage 27 company records.

Suggested next questions: Which of these have a director or PSC change in the last 12 months? / Which lenders hold the existing charges? / Who is the finance contact at the top ten?

DataGardener MCP is a Model Context Protocol server from DataGardener Solutions Limited that gives authorised AI agents and applications read-only access to connected UK company, financial, people, ownership, lending, procurement, public-spend, property and economic intelligence, with source and date metadata on records. Verified 2026-10-03.

Existing customer proof

Proven intelligence. A new way to access it.

These organisations use DataGardener intelligence today. DataGardener MCP is a new way to put the same intelligence in front of an AI agent.

Financial services

  • Allica Bank
  • Big Society Capital

Government and public sector

  • Oxfordshire County Council
  • Transport for London
  • Greater London Authority
  • City of London
  • Islington Council
  • NHS
  • The Dudley Group NHS Foundation Trust

Enterprise

  • Bibby Marine
  • Unilever
  • GSK
  • WSP
  • Serco
  • Sue Ryder
  • MSDUK

Universities and research

  • Aston University
  • Bournemouth University
  • Cardiff University

Logos indicate organisations using the DataGardener platform, as shown on datagardener.com. They do not indicate use of DataGardener MCP.

Trust is the foundation

AI is new. The intelligence behind it does not have to be unproven.

Explore the Trust Centre

The idea

Artificial Intelligence is powerful.Actual Intelligence makes it useful.

AI models can reason. Agents can act. Enterprise decisions still require intelligence that is trusted, connected, current and source-aware. DataGardener supplies that Actual Intelligence.

  • AI models reason

    They weigh conditions, plan steps and explain conclusions.

  • Agents act

    They choose tools, call them, read the results and continue.

  • DataGardener provides the evidence

    Connected UK business intelligence, with the source and date on every record.

The question is no longer

“Where is the data?”

The question is simply

the question.

“Which of our top 50 suppliers changed ownership this year, and what do their latest accounts show?”

search_pscsearch_gs02search_ukgaap
SourcePSC register; group structure; filed accountsDateTo Oct 2026ModuleOwnership, FinancialUsage50 records

One connection. Many intelligence layers.

Twelve layers, one connected intelligence system.

Company, financial, people, ownership, risk, lending, procurement, spend, property, contacts, economy and trade. Joined on registration number, entitled per customer, queried by the agent only when the question needs them.

DataGardener MCP

One connection. Twelve layers.

  • Company Intelligence

    Core profile of every UK registered company: identity, status, SIC codes, addresses, incorporation and filing history, group membership.

    Sample fields
    companyRegistrationNumbercompanyStatussicCode07incorporationDateregisteredAddress
    Tools
    search_companiesexecute_querysearch_group_structuresearch_gs02
    Sources
    Companies House; DataGardener company index
    Sample question
    “Which engineering SMEs in the West Midlands were incorporated in the last ten years and are still active?”
  • Financial Intelligence

    Filed accounts across multiple years: turnover, profit, balance sheet, cash, employees, growth and derived ratios, drawn from statutory filings and UK GAAP tagged accounts.

    Sample fields
    turnoverprofitAfterTaxnetAssetsnumberOfEmployeesyearEndDate
    Tools
    search_ukgaapexecute_ukgaap_querysearch_statutory_accountssearch_accounts_data
    Sources
    Companies House iXBRL accounts; DataGardener accounts data index
    Sample question
    “Show manufacturing companies in Birmingham whose turnover grew but whose cash position weakened over the last two filings.”
  • People and Directors

    Current and resigned officers, appointments across companies, director networks and recent changes.

    Sample fields
    officerNameroleappointedOnresignedOnotherAppointments
    Tools
    execute_query
    Sources
    Companies House officers
    Sample question
    “Which suppliers on this list have had a director resignation in the last 90 days?”
  • Ownership

    Persons with significant control, shareholdings, corporate group structures, parents and ultimate parents, and investor relationships.

    Sample fields
    pscNamenatureOfControlultimateParentshareholderNamenumberOfSharesIssued
    Tools
    search_pscsearch_shareholdingssearch_group_structuresearch_investor
    Sources
    Companies House PSC register; Confirmation statements; DataGardener group structure
    Sample question
    “Who ultimately owns this supplier, and has control changed since the contract was awarded?”
  • Risk

    Signals that a company's position is changing: charges, CCJs where available, late filings, status changes, insolvency events, and DataGardener risk indicators.

    Sample fields
    riskScorechargesOutstandinglateFilingstatusChangeinsolvencyEvent
    Tools
    execute_querysearch_dgsafe
    Sources
    Companies House; Registered charges; DataGardener risk indicators
    Sample question
    “Flag any portfolio companies with new charges or overdue accounts this quarter.”
  • Lending

    Registered charges, lender relationships, charge dates and satisfactions, funding events and refinancing windows derived from charge history and financial change.

    Sample fields
    chargeHolderchargeCreatedOnchargeSatisfiedOnlenderNamerefinancingWindow
    Tools
    execute_querysearch_dgsafe
    Sources
    Companies House charges register; DataGardener lending intelligence
    Sample question
    “Which companies in Greater Manchester have a charge created three to five years ago with no newer charge?”
  • Procurement

    UK public-sector contract notices and awards: buyers, suppliers, values, CPV codes, regions, status and SME suitability.

    Sample fields
    noticeTitlebuyerawardedSupplierawardedValuecpvCode
    Tools
    search_contractsexecute_procurement_query
    Sources
    Contracts Finder; Find a Tender
    Sample question
    “Which construction SMEs in the North West have won public-sector contracts in the last two years?”
  • Public Spend

    Published payment data from UK councils, combined authorities, NHS bodies and government departments, resolved to suppliers.

    Sample fields
    buyersupplieramountpaymentDatecategory
    Tools
    execute_procurement_query
    Sources
    Published payments over £500 and £25,000; DataGardener spend data
    Sample question
    “How much have councils in the West Midlands paid this supplier over the last financial year?”
  • Property and Land

    Company-owned titles from HM Land Registry's corporate ownership data, plus price-paid transactions for market context.

    Sample fields
    titleNumberproprietorNametenurepricePaiddateProprietorAdded
    Tools
    search_corporate_landsearch_land_registryexecute_land_registry_query
    Sources
    HM Land Registry corporate ownership; HM Land Registry price paid
    Sample question
    “Which freehold titles does this company own, and when were they acquired?”
  • Contacts

    Business contact details for decision-makers where available and permitted, including email, phone and professional profile links, for entitled customers.

    Sample fields
    contactNamejobTitlebusinessEmailphoneprofileUrl
    Tools
    execute_query
    Sources
    DataGardener contact data
    Sample question
    “Who is the finance director at each of the shortlisted companies?”
  • Local and Regional Economy

    Business population, sector mix, growth, incorporations and procurement activity aggregated by region, county, local authority, constituency or postcode district.

    Sample fields
    regionlocalAuthoritybusinessCountsectorMixgrowthIndicator
    Tools
    execute_queryexecute_accountsdata_query
    Sources
    DataGardener aggregations across company and procurement indexes
    Sample question
    “Show high-growth digital businesses in Croydon founded in the last five years.”
  • International Trade

    UK importers and exporters, trading partners and global firmographics for companies with an international footprint.

    Sample fields
    importerExporterFlagtradePartnerCountrycommodityGrouprevenueRangeemployeeRange
    Tools
    search_global_revenue_companiesexecute_global_revenue_query
    Sources
    HMRC importer and exporter data; DataGardener global revenue dataset
    Sample question
    “Which UK manufacturers export to Germany and have grown revenue in the last two years?”

Provenance as product design

Your agent should show its working.

Every DataGardener MCP record can carry the source it came from, the date of that source, the intelligence module that produced it, the company registration number it belongs to and the usage consumed. See the evidence behind the answer.

  • Source

    Which register, filing or dataset the fact came from

  • Source date

    The period or date of the underlying document

  • Intelligence module

    Which layer and tool produced it

  • Record identity

    The Companies House registration number

  • Usage

    What the call consumed, reported with the result

How provenance works

A real record, with its provenance

search_ukgaap

AUTOMOTIVE PRECISION ENGINEERING LIMITED

CRN12143071StatuslivePeriod01/09/2024 to 31/08/2025
Turnover
Not disclosed (small company filing)
Cash at bank
£6,662
Trade debtors
£284,748
Shareholders' funds
£59,199
Employees
11
SourceCompanies House filed accounts (iXBRL)DateRetrieved 3 October 2026ModuleUK GAAP financials indexUsage1 company record

Public Companies House filing retrieved through DataGardener MCP on 3 October 2026. Shown to illustrate provenance only.

Outcomes first

What becomes possible when an agent can investigate.

Lending

Find refinancing opportunities before the application arrives.

Lending intelligence that an agent can act on.

Explore Lending
Procurement

Investigate a supplier from every angle.

Supplier discovery and due diligence, with the evidence attached.

Explore Procurement
Economic development

Understand a local economy without weeks of research.

Business population, growth and sectors, by the geography you actually work in.

Explore Economic development
Market research

Turn a strategic question into an evidence-backed briefing.

Market sizing from filings, not estimates.

Explore Market research
AI and agent builders

Give your product UK business intelligence without rebuilding the data infrastructure.

The intelligence layer is operated. You build the product.

Explore AI and agent builders

Traditional workflow versus agent workflow

Remove the research between the question and the decision.

Before

  1. Question
  2. Multiple sources
  3. Exports
  4. Spreadsheets
  5. Manual filtering
  6. Analyst research
  7. Cross-checking
  8. Report

Eight hand-offs. Each one costs time and loses provenance along the way.

With DataGardener MCP

  1. Question
  2. AI agent
  3. DataGardener MCP
  4. Relevant intelligence tools
  5. Connected evidence
  6. Answer or action

One question. The agent performs the investigation and returns the answer with its evidence attached.

Research-time calculator

What is manual research costing you?

Enter your own numbers. Every assumption is editable, every output is an estimate.

Assumptions (edit any of them)

Analysts, relationship managers, procurement officers or researchers.

Supplier checks, borrower briefs, market lookups, due-diligence reports.

min

Time across databases, exports, spreadsheets and cross-checking.

£

Salary plus on-costs, per hour.

%

Your estimate. The agent does the gathering and cross-referencing; people keep the judgement.

Estimated outcome

Monthly research hours today
162
Hours potentially recovered per month
97
Annual capacity recovered
1,169 h
about 0.7 full-time equivalents
Estimated productivity value per year
£52,609

This is an estimate computed only from the assumptions you enter, using 4.33 weeks per month and 1,650 working hours per full-time equivalent. It does not include DataGardener MCP costs and is not a guarantee of savings. Validate the acceleration share with a pilot before relying on it.

Sector workflows

Go deeper by sector

Existing DataGardener evidence

Outcomes from the intelligence behind the MCP.

These are published DataGardener platform case studies. They show what the underlying intelligence does for customers. They are not MCP success stories.

All case studies
Existing DataGardener case study

Data-driven success for Allica Bank's SME strategy

Allica Bank · Financial services

Allica Bank used DataGardener to identify and reach a target population of around 500,000 UK SMEs for its relationship-banking proposition.

target SMEs identified
~500,000
monthly customer enquiries from day one, per Allica
Doubled
“The volume of customer inquiries coming through the door pretty much doubled every month from day one.”
Conrad Ford, Chief Product and Strategy Officer, Allica Bank
Outcome of the DataGardener platform and Lending Intelligence, not of DataGardener MCP. Source
Existing DataGardener case study

Building supplier diversity faster with a supplier intelligence platform

Unilever UK · Enterprise procurement

Unilever UK used DataGardener profiling and enrichment to identify diverse suppliers already in its network and to reduce the manual research involved.

diverse suppliers identified
100+
reduction in manual research time
80%
“With DataGardener's profiling and enrichment tools, we gained clarity on which suppliers in our network were diverse and could be supported further. The speed and accuracy have fundamentally improved how we approach supplier diversity.”
Procurement and Supplier Diversity Lead, Unilever UK
Outcome of the DataGardener Supplier Intelligence, not of DataGardener MCP. Source
Existing DataGardener case study

How a broker closed 37 invoice finance deals with Lending Intelligence

UK invoice finance broker (unnamed) · Commercial finance

A UK invoice finance broker used real-time charge signals and direct contacts to find companies with financing needs earlier.

invoice finance deals in 12 months
37
“Thanks to DataGardener's Lending Intelligence, we find companies with genuine financing needs earlier than ever. The real-time charge signals and direct contacts helped us close 37 deals in a year.”
UK invoice finance broker
Outcome of the DataGardener Lending Intelligence, not of DataGardener MCP. Source

Easy start

Four steps from connection to investigation.

No pipelines to build, no schemas to learn. Tested today with Claude (web, desktop and mobile, via Connectors); any MCP client that supports remote servers with OAuth can connect.

Connection guideBook an MCP demo
  1. 1

    Connect

    Add DataGardener MCP to a supported MCP environment as a remote server.

  2. 2

    Authenticate

    Authorise with OAuth. The connection is bound to the correct DataGardener entitlement.

  3. 3

    Ask

    Use natural language. No query syntax, no exports.

  4. 4

    Investigate

    The agent chooses the relevant DataGardener intelligence tools and shows its working.

Developers and OEM

DataGardener operates the intelligence layer. You build the experience.

Build UK business intelligence into your agent or product without building and maintaining the underlying data infrastructure. 34 task-level tools, OAuth authorisation, read-only by design, usage reported with every result.

Developer overviewOEM partnerships

What your team does not have to build

  • Source ingestion
  • Normalisation
  • Entity resolution
  • Business-data pipelines
  • Procurement aggregation
  • Company intelligence infrastructure
  • Continual source updates
// agent decides, DataGardener MCP answers
tool: search_companies
args: { query: "precision engineering", region: "west midlands",
        company_status: "live", size: 6 }
→ Found 10,000 matching companies (showing top 6)
→ each record: name, CRN, status, industry, location,
   size band, age; usage reported with the result

Questions buyers ask first

What is DataGardener MCP?

DataGardener MCP is a Model Context Protocol server from DataGardener Solutions Limited. It gives authorised AI agents and applications read-only access to connected UK company, financial, people, ownership, lending, procurement, public-spend, property and economic intelligence, with the source and date on records.

Which AI agents can use DataGardener MCP?

Any MCP client that supports remote servers with OAuth authorisation can connect. DataGardener MCP has been used with Claude through Connectors. Other clients are not yet individually tested by DataGardener.

Is DataGardener MCP a replacement for the DataGardener platform or API?

No. The platform is for people working interactively, the Company Data API is for deterministic integrations, and DataGardener MCP is for AI agents that decide which intelligence tools to call. All three draw on the same intelligence.

Do existing DataGardener customers already use the MCP?

The customer logos on this site show organisations that use DataGardener intelligence through the platform. They do not indicate MCP adoption. MCP access is added to a customer's entitlement.

How is DataGardener MCP secured?

Connections are authorised with OAuth, so identity travels with the connection rather than with a shared key. Tools are read-only and the server rejects writes. Each connection sees only the modules its DataGardener entitlement allows. DataGardener holds ISO/IEC 27001 and Cyber Essentials certification.

What does an answer look like?

A short answer, the matching companies or aggregates, and the evidence behind them: source, source date, intelligence module, company registration number and the usage consumed. The agent shows its working.

Your agents already know how to think.Give them something worth knowing.

Built on DataGardener intelligence already used by leading UK organisations.

Book an MCP demoDiscuss an agent use case