Home / AI Agents
The agent roster

Sixteen agents. One data spine.

Each agent owns a job end to end — fetching its own data, running its own scientific or calculation standard, and writing back to a shared ledger the others can query. Together they cover data collection, emissions inventory, risk and decarbonisation modelling, and reporting.

3 · Data collection 3 · Inventory & accounting 7 · Risk & decarbonisation modelling 3 · Reporting & governance
01 · Data collection

Feeding every model below, automatically.

COLLECT Data Collection Agent

Fetches data from any connected source — ERP, finance, utility, spend and metered-data systems — across your organisation, assets, services and supply chain, or across an entire portfolio.

COLLECT Supply Chain Agent

Runs supplier-level data requests in plain language and reconciles what comes back straight into the inventory — replacing spend-based estimates with real activity data as coverage grows.

COLLECT Gap Analyzer Agent

Continuously scans the inventory and risk models for missing or stale data and flags it before it becomes a reporting problem.

02 · Inventory & accounting

Emissions accounting, simplified — every activity accounted for.

CALCULATE Emissions Ledger · S1 · S2 · S3

Builds and reconciles scope 1, 2 and 3 accounting from ERP, utility, spend and supplier data as it lands, reconciled to the GHG Protocol Corporate Standard and Corporate Value Chain (Scope 3) Standard.

CALCULATE Life Cycle Agent

Product and material footprints across cradle-to-gate or cradle-to-grave boundaries, aligned to ISO 14040/14044, built on the same activity data as the corporate inventory.

REPORT Offsets Ledger Agent

Credit inventory, quality screening against project-level registries, and retirement and claims tracking with a full audit trail.

03 · Risk & decarbonisation modelling

Turning the inventory into a plan.

MODEL Transition Risk Agent · RISK-T

Scenario modelling for policy, carbon price, technology and demand shifts across the organisation, its assets and supply chain — translated into financial impact and ranked resilience actions, not left as a bare exposure score.

MODEL Physical Risk Agent · RISK-P

Exposure to heat, flood, drought, wildfire and cyclone, built on downscaled climate models for a more accurate, localised read than a standard global model — modelled under multiple forward pathways.

MODEL Scenario Agent

Builds and maintains the climate-science-backed scenarios that RISK-T, RISK-P, MACC and TAC all run against, so every risk and abatement output is working from one consistent set of assumptions.

MODEL Entity Risk Aggregation Agent

Rolls asset- and organisation-level risk up to the counterparty or portfolio level — so a lender, insurer or asset owner can see an entire book at once, not one exposure at a time.

MODEL Abatement Agent · MACC

Marginal abatement cost curves that rank every reduction lever your business actually has by cost per tonne and available volume, refreshed as energy and carbon prices move.

MODEL Adaptation Agent

Prices specific resilience measures — flood barriers, HVAC upgrades, structural hardening — against avoided-loss value, closing the loop from “here’s your risk” to “here’s what it’s worth to fix.”

MODEL Trajectory Agent · TAC

Overlays your net-zero trajectory onto the abatement curve and flags exactly where committed reductions fall short of target.

04 · Reporting & governance

Clear answers, and filings that write themselves.

REPORT Disclosure Agent · DISC

Drafts and maintains filings for IFRS S2, CSRD/ESRS, TCFD, PCAF and CDP from one underlying data model, mapped to jurisdictional overlays, with a readiness check that can be run any week of the year.

MODEL Financial Impact Agent

Translates the ledger, risk exposures and abatement curve into balance sheet and P&L terms — climate value-at-risk, stranded-asset exposure, transition capex — at the entity or portfolio level.

REPORT Climate Copilot

A plain-language interface over the whole data spine — ask where a number came from, what’s driving a change this quarter, or which action closes the largest exposure — and get an answer with source records cited.

See the agents run on your own data.