Uzbekistan has already built e-gov. What is the next layer

What is already done

Uzbekistan has built what the industry calls an e-government foundation. A unified MyGov portal with hundreds of services. OneID identification. Electronic signatures. Registries of population, real estate, transport, business. A mobile app with significant adult-population coverage. Integration with banking and payment apps (dgov.uz).

This is not “the start of digitisation”. It is already a working baseline infrastructure. By public indicators internet coverage is 99.5%, user penetration 94.2% (dgov.uz), 98% of populated areas have mobile internet (gov.uz).

In this reality, selling the state “let’s build another portal” or “let’s automate one more service” misses the moment. The basic portal exists. Basic automation exists. Further value comes not from adding features, but from a structural shift in how the state works with the citizen and the business.

Where the shift is

In the current model the citizen and business initiate contact with the state. They open the portal, pick a service, fill in a form, wait. A reactive model. It works, but it requires the user to know they have an obligation, find the service and submit a request.

The next layer is a state that initiates contact first. Knows your document is approaching expiry. Knows your business has a new regulatory obligation. Knows you went through an event (birth, move, opening a company) that usually triggers a package of actions, and proposes them as one route.

This is not “more functions”. It is a different operating model of the state. The citizen does not have to search — the state comes first. The business does not have to monitor — the state notifies.

What this layer requires

Several structural components missing in most e-gov systems.

Personal Government Companion. Not just a personal cabinet but an interface that knows the user’s context — where they are, what obligations are coming up, what to do, what the state already knows so it does not have to be repeated. A separate article covers this component.

Life-Event Orchestration. When a life event happens (birth, marriage, move, opening a business, retirement), instead of the citizen walking through 5-10 services, the state links them into a single route. One pass, everything done.

Government Memory Layer. If a citizen has already proven something to the state (showed passport, confirmed income), they should not have to do it again on every subsequent interaction. The state remembers context.

Public Trust Layer. The citizen sees who has accessed their data and why. They can explicitly grant or revoke consent. They can challenge a decision — and see what it was based on.

AI Case Officer. An assistant for the civil servant, not a replacement. AI prepares a draft decision, the human checks and finalises. This cuts processing time by multiples while preserving accountability.

Each of these is a separate product. Together they form what we call Intelligent Government — the next layer above the existing e-gov.

What no longer works in the current model

Despite a working infrastructure, several patterns remain painful.

Repeated requests. The citizen provides the same information several times — across services, across agencies. The state already has the information but requests it again because there is no shared contextual layer.

Silent obligations. The citizen or business has an obligation they do not know about (renewal date, new regulatory requirement). They find out only after the fine. This creates the feeling that “the state catches, it does not help”.

Unintelligible refusals. A service is refused. The reason — a few lines of legal text. The citizen does not understand what to fix. They reapply, the refusal repeats.

Cross-agency channels. The citizen contacts one agency, but the issue requires resolution at another. The state does not route automatically — the citizen has to walk it themselves.

No proactive notification on relevant changes. The law changed, the rule changed, the citizen or business has a new obligation — they find out from acquaintances or the news, not from the state.

What the state in 2035 looks like after this layer

In ten years a digital state that has built this layer looks like this. The citizen rarely opens the portal on their own initiative — most interactions arrive as a notification from the state with a one-click “accept” or “decline”. The business knows its obligations 30-60 days in advance and sees an action checklist. Life events (marriage, move, birth, retirement) are handled as one route, not 10 separate services.

The civil servant spends most of their time on non-standard cases. Standard cases are processed by AI with human verification. Not “AI takes decisions for humans” — “AI removes the routine, the human focuses on the complex”.

Trust grows because the citizen sees that the state knows the context, explains decisions, does not require the same thing repeatedly.

What needs to start in the next 24 months

Not “everything at once”. Phased.

First 6 months. Strategic clarity. The digitalisation council formulates which 3 of the 5 components above are top priority on a 5-year horizon. Not all at once — a choice.

Months 7-12. Architectural foundation. Event Graph (the flow of citizen and business events), Government Memory Layer (shared context), Identity and Consent Layer (consent management).

Months 13-18. First pilots. One pilot on Life-Event Orchestration (for example, child birth). One pilot on Personal Government Companion. One pilot on AI Case Officer in one agency.

Months 19-24. Expansion. Pilots become production. A second agency joins. Customer feedback and refinement.

By two years in — visible results on 2-3 use cases. By 5 years — critical mass. By 10 years — the next layer as standard.

When the topic is premature

If the basic infrastructure is unstable (the portal often fails, OneID has verification problems, registries diverge), stabilise the base first. The intelligent layer over a broken foundation makes problems worse.

If cross-agency interaction in the current reality is weak (each agency protects its data, does not share), Life-Event Orchestration will not start. Data governance and agreements first.

If regulation around personal data and AI decisions is unsettled, AI Case Officer and Predictive Government operate in a legal grey zone. Legal clarity first.

If the management team has no role accountable for “the next layer” as a product (a Chief Digital Officer for the state), the initiative dilutes between agencies.

If the budget is constrained on a 1-2 year ROI, while Intelligent Government has a 5-10 year ROI, the initiative will not receive sustained funding.

Discussion at the digitalisation council level

Which 3 of the 5 components are priorities for our country on a 5-year horizon? All 5 simultaneously — unrealistic.

What is the current maturity of the basic infrastructure? Is it ready to carry the next layer?

Who owns the next layer as a product? If distributed across agencies — fragmentation.

What regulatory work is needed for AI Case Officer and Trust Layer in the next 12 months?

What 5-year investment horizon is acceptable to the cabinet of ministers and the budget process?

How SamaraliSoft can help

Intelligent Government 2035 Vision Sprint — diagnostic of current e-gov maturity, selection of priority 3 of 5 components for the next layer, architectural blueprint for Event Graph and Memory Layer, regulatory engagement plan, and a roadmap for the first 24 months with pilots on 2-3 use cases.

Related reading

• /en/insights/government-personal-companion/ — Personal Government Companion
• /en/insights/government-portal-is-yesterday/ — why “another portal” is yesterday
• /en/insights/government-life-event-orchestration/ — life-event orchestration
• /en/insights/government-public-trust-layer/ — trust layer

Sources

• Digital Government Portal — state of digital infrastructure UZ
• Ministry of Digital Technologies — telecom and digital UZ
• UzDaily — digital and telecom market UZ
• Central Bank of Uzbekistan — biometric verification for digital banking