Mongolia's Ministry of Economy and Development embedded an AI assistant, 'AI Туслах', in its planning platform, helping planners across 39 ministries define KPIs aligned with the 2026-2030 Development Guidelines. Still pilot-stage; no outcome evaluation published yet.
39 ministries/agencies
Ministries and agencies compiling development-plan KPIs (2023 planning cycle)
34,000+ rows
Rows of KPI data compiled manually before the platform (2023)
38 outcomes
Green outcomes in the Medium-Term Development Plan (2026-2030)
60 KPIs
KPIs in the Medium-Term Development Plan (2026-2030)
76 indicators
SDG-aligned indicators in the plan (2026-2030)
Details
Maturity
Pilot
Promoter
Ministry of Economy and Development (Mongolia), with UNDP Mongolia / UN PAGE
Period
2025-2026 (pilot)
Keywords
public administration, development planning, monitoring & evaluation
Context
Mongolia's public administration faces high staff turnover and limited monitoring-and-evaluation capacity. Previously, ministries submitted development plans manually: 39 ministries and agencies compiled over 34,000 rows of KPI data across spreadsheets for the 2023 planning cycle, a process prone to duplication and inconsistency.
Activities
The Ministry of Economy and Development (MED), supported by UNDP's Strengthening Systems Approach for SDGs in Mongolia (SASiM) project and the UN Partnership for Action on Green Economy (PAGE), built a digital planning platform (plan.med.gov.mn) with an embedded AI assistant, 'AI Туслах'. When a planner enters a project, the assistant suggests relevant KPIs — such as installed capacity, construction progress, electricity generation, budget execution, jobs created — with units, baselines and data sources, and validation tools automatically flag duplicate entries, unverified agency names and indicator/budget inconsistencies.
Results
The tool links planning entries directly to Mongolia's Medium-Term Development Plan (2026-2030) and its 38 green outcomes, 60 KPIs and 76 SDG-aligned indicators. As of the source's publication date (April 2026), it remained in pilot phase with no published data on adoption rates, KPI quality improvement, or planning-cycle time savings — the evidence available is descriptive/process-level, not outcome-evaluated.
Implementation
Indicative cost
Medium (€50k–€500k) — Not quantified in sources; built as a government digital platform under existing UNDP SASiM/PAGE technical support rather than a described separate budget line.
Time to results
Medium (1–3 years) — Platform in pilot phase 2025-2026 as of the source's April 2026 publication date, tied to the 2026-2030 Medium-Term Development Plan cycle.
Staffing & skills
Planners across 39 ministries and agencies use the platform to enter projects and KPIs, Technical support from UNDP's SASiM project and UN PAGE
Conditions for success
Built-in validation tools that automatically flag duplicate entries, unverified agency names and indicator/budget inconsistencies
Direct linkage of planning entries to the binding Medium-Term Development Plan (2026-2030) targets
Common failure modes
Prior manual, spreadsheet-based planning process across 39 ministries was prone to duplication and inconsistency in KPI data, the problem the platform was built to address
High staff turnover and limited monitoring-and-evaluation capacity in Mongolia's public administration are named risk factors for sustained platform use
Commonly funded by
National / regional programmesPhilanthropic / foundation funding
Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.
Do you run this practice?
Claim it —
verified implementers get a public contact pathway and can propose corrections.
Data sources
Where this practice's information was retrieved from, and when.
Germany's Federal Statistical Office trained a machine-learning model to improve the accuracy of mandatory compliance-cost ('Erfüllungsaufwand') estimates for draft regulations …