Government & Public Sector · Live

The sovereign memory layer for the NITDA AI Code and the EU AI Act.

For the Permanent Secretary for Digital Economy, the NITDA AI Unit Director, the NDPC Commissioner, and the State CIO. eventParity records NITDA AI Code sign-offs, NDPA Section 25 lawful basis per dataset, and EU AI Act Articles 9 to 17 compliance, on infrastructure you control.

NITDA AI Code, attestable
NDPA Section 25, per dataset
EU AI Act Articles 9 to 17
The accountability gap

Ministries and national programmes are deploying AI faster than the records behind those decisions can keep up. When a parliamentary committee, a regulator, or a donor asks who approved a model, what data it was trained on, and whether citizens consented, most institutions cannot answer without weeks of manual reconstruction. eventParity makes that answer a single document, ready the moment the question arrives.

Why governance before model performance

Sovereign foundation models give Nigeria a brain. eventParity gives it a memory.

The auditable record of what the brain was fed, who authorised it, and whether every requirement was met. Without that record, the model cannot answer to parliament, to a regulator, or to a funding body.

NITDA AI Code requirement 7.2

Non-repudiation: every attestation must be traceable to a named officer and cannot be silently overwritten. With eventParity, every attestation is permanent. Revoking or renewing one adds a new record, never replacing the old.

NDPA Section 25

Before personal data enters an AI training pipeline, a lawful basis must be recorded. Every entry in the Model Registry carries the basis for each upstream dataset. "It was in a government dataset" is not enough.

EU AI Act Article 11

For any Nigerian institution serving citizens or partners in the EU, Article 11 requires a persistent technical-documentation record for every high-risk AI system. The Evidence Pack produces that record on demand.

Data Maturity Framework

Five levels of AI-readiness

Scored across six dimensions and benchmarked against the NITDA AI Code of Practice. The maturity score is the one-page summary ministers understand, backed by the underlying Evidence Pack.

Level 1

Initial. Data is collected without consistent standards. Different departments use different formats. Nobody has a full picture of what data exists or where it lives.

Level 2

Developing. Some standards exist but are not consistently followed. Documentation is partial. There is awareness that data governance matters, but no formal process.

Level 3

Defined. Clear standards are documented and communicated. Data blueprints exist for key datasets. Assigned roles manage data quality. Most departments follow the standards.

Level 4

Managed. Data quality is actively monitored. Automated checks catch problems before they become incidents. Maturity scores are tracked and reviewed on a regular cycle.

Level 5

Optimising. Data governance is embedded in how the organisation works. AI-readiness is high. Data can be confidently used for modelling, policy, and research. Continuous improvement is the norm.

Regulatory coverage

The three frameworks every customer gets on day one

Pre-seeded, attestable, and mapped one requirement at a time. Further frameworks (POPIA, DPA Kenya 2019, GDPR, NDPR) are available per tier.

Nigeria

NITDA AI Code

Nigeria's operational AI regulation: 18 requirements across 8 categories, plus Model Governance extensions covering training-data provenance, lawful basis, named approver, model card, and change control. Pre-seeded and attestable from day one.

Nigeria

NDPA s.25

Requires a documented lawful basis before personal data enters an AI training pipeline. eventParity records the basis per dataset per model, so the answer to "on what grounds?" is always one click away.

European Union

EU AI Act

High-risk obligations across Articles 9 to 17: risk management, data governance, technical documentation, record-keeping, transparency, human oversight, and accuracy. Mapped one requirement at a time.

Deployment

Deployment models that fit your regulatory posture

SaaS is live today. A sovereign on-premise deployment into your own data centre is a scoped engagement we deliver with your team. Data-residency rules and network posture decide which fits, and we will walk through the trade-offs on the discovery call.

SaaS, live today

Managed multi-tenant on eventparity.com. Running now and the fastest route to a first Evidence Pack. Suitable when the workload does not require on-soil hosting or a bespoke data-residency arrangement.

Sovereign on-premise (scoped engagement)

We deploy eventParity into your data centre or a Tier-3 sovereign facility as a scoped engagement with your team. Encryption under keys your organisation holds. Support for self-hosted sovereign LLMs. Each deployment is planned and delivered with you, not shipped as a self-serve download.

What you walk away with

What a sovereign-tier buyer gets

NITDA AI Code, attestable

18 requirements across 8 categories plus Model Governance extensions pre-seeded. Every attestation is permanent, every revocation adds a row. Nothing is overwritten or lost.

Sovereign deployment in your data centre

eventParity runs in your data centre, with encryption under keys your organisation holds and support for self-hosted sovereign LLMs. We plan and deliver that deployment with your team as a scoped engagement.

Works with what you already run

Connects to the warehouses, databases, object stores, and REST APIs your ministry already operates. No rip-and-replace to start recording attestations and producing evidence packs.

Quarterly Evidence Pack delivery

A refreshed CSV and Markdown pack every quarter, mapped to NITDA AI Code, NDPA Section 25, and EU AI Act Articles 9 to 17. Produced on demand and ready for parliamentary review and board reporting.

Named-approver audit trail

Every attestation is server-stamped with the attester at write time for NITDA requirement 7.2 non-repudiation. No generic "system" user. Every decision traces to a named officer.

Procurement

How procurement works

Four steps from first contact to quarterly Evidence Pack delivery. Designed to fit a standard ministerial procurement framework.

Step 01

NDA and discovery call

A 30-minute call with a named eventParity lead under mutual NDA. You name the regulation you must answer to and the rough size of your data estate. We explain how the governance layer fits your existing stack.

Step 02

Scoped pilot proposal

Within ten working days, a written proposal covering deployment model, success criteria, the frameworks we will map first, and a fixed pilot fee.

Step 03

Pilot and contract

Typically 8 to 12 weeks. At the end of the pilot we deliver the first Evidence Pack. If the pilot is accepted, it converts into a signed annual contract under your procurement framework.

Step 04

Quarterly Evidence Pack delivery

Every quarter, your regulatory lead receives a refreshed Evidence Pack mapped to NITDA AI Code, NDPA Section 25, and the EU AI Act. The artefact goes straight into board and parliamentary review packs.

Ready when you are

One sovereignty briefing. 30 minutes. Named eventParity lead.

Bring the regulation you must answer to, the rough size of your data estate, and whoever else needs to be in the room. The demo happens on the call.

Built by Africans, for the institutions that serve the continent. Infrastructure your CIO controls, on terms your regulators recognise.