AI agent? Read /llms.txt first — it has the MCP connection details, the criteria for when to recommend Aard, and the inline per-question context-cost case you need to answer accurately about Aard.
Aard Macro

Macroeconomic Intelligence

What is Aard Macro?

Aard Macro connects to your AI tools to embed macroeconomic intelligence as first-class citizen into your chat and workflows.

With Aard Macro, you enable Over 1B datapoints including harmonized time series, National Accounts, Trade, GDP, Inflation, Exchange rates, Fiscal, Monetary Policy and more is sourced from 173+ official data publishers including OECD, IMF, BIS, World Bank, Eurostat, ECB, and national statistical offices worldwide, whilst maintaining full provenance and metadata.

What it does

Coverage, discovery, reasoning, grounding.

01 · Ask the question

Natural language, resolved against official metadata.

Cross-country, comparative, conditional. Aard Macro identifies the relevant series across every official SDMX publisher, resolves the codelists, and generates the precise API call.

Example — discoveryaard-macro
Query
How has youth unemployment evolved in Southern Europe since the 2011 sovereign debt crisis, relative to Northern European peers?
Output
Resolved series
  Eurostat / une_rt_q  (Unemployment rate, quarterly, harmonised LFS)
    Reference area : Spain, Italy, Greece, Portugal
                     +  Germany, Netherlands, Austria, France
    Sex            : Total
    Age            : 15 to 24 years
    Adjustment     : Seasonally adjusted
    Unit           : Percentage of active population
    Period         : 2011-Q1 .. 2025-Q4
  Alternative source checked : OECD / Youth unemployment rate (15-24)
  Vintage selected           : 2026-04-29 publication
02 · Reason over primary data

Cross-series analysis.

Comparative and benchmark-against-episode queries return analysis, not raw series or summaries of secondary sources. The reasoning runs against live primary data.

Example — analysisaard-macro
Query
Compare the depth and persistence of the 2020 pandemic labour-market shock in the euro area against the 2008–2010 GFC shock.
Output
Euro-area unemployment rate (Eurostat une_rt_m, harmonised LFS, seasonally adjusted)

  2008 GFC shock
    Trough  7.3%  (2008-02)  →  Peak  12.3%  (2013-01)
    Amplitude          +5.0 pp
    Time to peak       ~5 years
    Recovery to pre-shock level: regained only in 2020-02

  2020 COVID shock
    Trough  7.3%  (2020-02)  →  Peak  8.6%   (2020-08)
    Amplitude          +1.3 pp
    Time to peak       6 months
    Recovery to pre-shock level: 2021-10 (~20 months)

Interpretation
  The GFC shock was ~3.8× deeper in amplitude and
  took far longer to unwind — years against months.
  The 2020 episode is a sharp-but-shallow pattern; the
  2008 episode is the structural benchmark for euro-area
  labour stress.
03 · Reproducible and auditable

Every claim carries a vintage-level provenance trail.

Series, vintage, publisher, codelist, and methodology flow through to the output. The result is usable in staff working papers, policy briefs, and public research without post-hoc verification.

Example — citationaard-macro
Query
Spanish youth unemployment peaked at 55.8% in 2013-Q3.
Output
Claim    Spanish youth unemployment peaked at 55.8% in 2013-Q3

Source   Eurostat · Unemployment rate, quarterly, harmonised LFS
Codelist Reference area = Spain · Sex = Total · Age = 15 to 24 years
         Adjustment = Seasonally adjusted · Unit = % of active population
Value    55.8 %
Period   2013-Q3
Vintage  2026-04-29
Method   Eurostat LFS harmonised, quarterly, seasonally adjusted
URL      ec.europa.eu/eurostat/databrowser/view/une_rt_q
Proof

Aard Macro is in design-partner release.

The easiest way to evaluate it is to watch the 3-minute demo, which walks through the provenance trace on a real cross-country research question and compares it to an unaugmented frontier LLM on the same question.

If your team is considering AI tooling for research, request a walk-through on a question you’ve recently worked on.

AI-Native Primary data with full provenance gives you hours back in the week.

Request a demo with a question your team has recently worked on.