EDIS
Economic Development Intelligence Infrastructure
From data to decisions.
EDIS connects economic models, sector planning systems, public finance, GIS, simulation, optimization, forecasting, microsimulation and AI agents into one integrated decision-intelligence platform for governments and institutions.
Everything an allocation decision needs, in one system
An integrated economic development intelligence platform
EDIS connects macroeconomic models, fiscal models, sector models, public finance, public investment appraisal, GIS, microsimulation, climate modelling, optimization, forecasting, AI agents and reporting — so that a question entered once is answered across all of them, on one consistent basis. How EDIS works →
AI agents
Twenty-eight registered agents interpret a policy question, select the models that can answer it, and assemble the workflow. Agents orchestrate; they do not guess numbers.
Browse the agent registry →Models and simulation
FINEX, MTEF, CDCGE, UCSN, PIAS and fifteen sector and cross-cutting model services produce every quantitative estimate, forecast, optimization and appraisal.
Open the model explorer →GIS and spatial planning
Population, settlements, roads, facilities, grids, hazard and projects. EDIS answers not only what is needed but where it is needed, and who it reaches.
See the maps workspace →Reports and analytics
Executive summaries, policy briefs, technical reports and investment appraisals, each carrying the model used, the data used, the scenario, the assumptions and a run ID.
See report formats →A policy question becomes a routed analytical workflow
EDIS never sends a policy question to a language model and returns the reply. The question is classified, the models are resolved, the data and permissions are checked, and only then is a workflow executed — in this order, every time.
One question, ten models, one traceable answer
“What would happen to GDP, fiscal balances, employment, poverty and sector investment requirements if public investment in renewable energy increased over five years?”
No single model answers that question. EDIS runs the chain, shows each model as it completes, and stops rather than guessing when a model does not converge or a dataset is missing. That cross-model workflow is what separates EDIS from a chatbot or a standalone model.
Twenty-one model services, one framework
Every quantitative result in EDIS comes from one of these model services. They share a costing spine, a spatial layer and an economy-wide closure, so a health question and an energy question can be compared on the same basis.
Macroeconomics and public finance
- FINEX
- MTEF
- Budget Output Engine
- CDCGE
The envelope, the ceilings, what the allocation buys and the economy-wide closure.
Public investment and results
- PIAS
- UCSN
- MEL
Appraisal and ranking, unit costs and service norms, and the results framework behind reporting.
Sector models
- Agriculture
- Health
- Education
- Energy
- Transport
- Trade
- Industry
- Minerals
- Water
- ICT
- Forestry
- Fisheries
- Tourism
Sector planning linked to the budget and to the macroeconomic closure rather than standing alone.
Distributional analysis
- Poverty microsimulation
- Household analysis
- Income distribution
Who gains, who loses and by how much, on the country’s own survey data.
Climate and environment
- Climate risk
- NDC analysis
- Adaptation
- Mitigation
- Climate fiscal risk
Climate scenarios carried through to fiscal and sector consequences, not reported separately.
Sixteen sectors, one framework
Each sector sheet sets out the questions it answers, the models behind it and how it connects to the macroeconomic, fiscal and national planning systems.
Agents make the models usable. They do not replace them.
The AI layer and the analytical layer are separated by design, and the separation is enforced in the architecture rather than merely described in the marketing. An agent routes, orchestrates, validates and explains; every number comes from a model.
- Understand — interpret the policy question and its scope
- Identify data — find the datasets that can support an answer
- Identify models — select the model services that accept the question
- Construct the workflow — order the chain and its dependencies
- Execute and coordinate — queue, run and monitor the models
- Validate — convergence, plausibility and consistency checks
- Interpret — explain what the results mean and what they do not
- Report — assemble the decision report with full provenance
The AI layer
- Understands the question
- Routes it to registered agents
- Orchestrates the model workflow
- Explains what the results mean
- Writes the report around them
The analytical layer
- Calculates and forecasts
- Simulates scenarios
- Optimizes allocations
- Maps and analyses space
- Validates every result
| Question | Sector | Access |
|---|
A registered question library: every question class carries its agent, its workflow, its models, its data requirements and its validation status.
The institutions that allocate resources
Each sheet sets out the questions the institution has to answer, the models behind them, the outputs it gets and how a deployment is licensed.
Deploy EDIS across your institution
An institution can begin with one analyst and grow to a national platform without changing tools or rebuilding its data integrations. Implementation covers country configuration, model calibration, data integration, institutional setup, user roles, API integration, training, testing, deployment and support.
- 01IndividualOne economist on EDIS Professional.
- 02Analytical teamA unit on EDIS Team, with shared workspaces and an audit trail.
- 03MinistryOne ministry on its own configuration, models and data.
- 04Government enterpriseSeveral institutions on one governed estate.
- 05National platformWhole-of-government access and the national data connection.
- 06SovereignA sovereign configuration agreed with the government.
From individual analysis to national decision infrastructure
Individual plans are paid online. Institutional and government deployments are contracted, because what is licensed is a configured national estate rather than a number of logins.
Subscription
The platform licence: users, models, countries, data and governance. The core of the business.
Plans and pricing →Implementation
Country configuration, calibration, data integration, institutional setup and acceptance testing.
Implementation →Academy training
Onboarding, analyst certification, administrator training and government cohorts that support adoption.
Training pricing →Custom development
New models, custom agents, bespoke integrations and dedicated environments, scoped and priced separately.
Custom development →Explore, use, adopt, deploy, integrate, scale
Each stage is a real licence with real entitlements, and each one leads to the next only when the institution has a reason to go there.
Academy training and certification support every stage of adoption. Training inside a deployment →
What a minister actually receives
A run does not end in a table of numbers. It ends in a document structured the way a decision has to be argued — the answer first, the assumptions next, the consequences in order, and the provenance at the end so the whole thing can be checked.
See report formats Example outputs
- Data source
- Utopia National Household Survey
- Data year
- 2019/20
- Country
- Utopia
- Model
- CDCGE
- Model version
- 5.1
- Scenario
- AGR-CAPEX-1T
- Run ID
- EDIS-DEMO-2026-0413
- Validation
- Validated
Every result carries a provenance record. If a figure cannot be traced to a model version, a dataset and a run, EDIS does not display it.
Structure of a decision report
- Policy question
- Executive answer
- Scenario assumptions
- Macroeconomic results
- Fiscal results
- Sector results
- Employment
- Poverty and distribution
- GIS maps
- Charts
- Investment requirements
- Risks
- Policy options
- Model and data provenance
What makes it different
Integrated
Multiple models and systems in one platform, rather than a folder of disconnected spreadsheets and tools.
Evidence-based
Results are produced by data and analytical models, each traceable to a dataset and a model version.
Cross-sector
Links macroeconomics, fiscal policy and sector planning, so sector plans are costed against the envelope.
Spatial
GIS and geographic analysis are part of the answer: where it is needed and who it reaches.
Scenario-based
Policy simulation rather than static reporting: what changes, under which assumptions, by when.
AI-enabled
Agents help users navigate and orchestrate analytical workflows; the models still produce the numbers.
Government-oriented
Built around real planning, budgeting, investment and policy workflows, and procured the way governments procure.
Failure is shown
Non-converged runs and missing data are displayed, not hidden, because hiding them would misrepresent the evidence.
Training that supports EDIS adoption
EDIS Academy provides specialised training and certification for economists, planners, policymakers and analysts who use EDIS and its underlying economic, fiscal, sector and planning models. Courses are built around the platform itself: learn, practise, certify, deploy.
Institutional and government deployments can include user onboarding, model training, analyst certification, administrator training, refresher courses, government cohort training and advanced model courses — so that the institution can run its own analysis after the implementation team has gone.
See EDIS answer a question your institution has to decide
We run a demonstration against your own policy question, in your country, with the data and model limitations stated.
Request an EDIS demo Open the public demonstration Talk to our team