The digital transformation of public service delivery is reshaping sectors worldwide, with education emerging as one of the most profoundly impacted domains. Increasingly, digital technologies are being adopted not only to enhance operational efficiency but also to strengthen data security and promote inclusivity within education systems [1]-[3]. This transformation reflects a shift from the mere digitisation of processes to a fundamental rethinking of how educational data are collected, managed, and utilised to support evidence-based decision-making. Within this evolving context, the integration of digital platforms into educational assessment has become a strategic priority, as it supports the development of systems that are equitable, transparent, and resilient in meeting the demands of twenty-first-century education [4].

In Tanzania, however, the educational assessment system continues to rely heavily on manual, paper-based, and fragmented processes. Although initial efforts have been made to digitise specific functions, such as student registration, the broader assessment lifecycle from candidate verification to results processing and certification remains inefficient and vulnerable to data inaccuracies. These inefficiencies disproportionately affect rural and underserved regions, where access to timely and reliable educational services is limited [5]. Furthermore, the absence of interoperability between education systems and other key government sectors, including civil registration, health, and social welfare, constrains seamless data verification and information sharing. This fragmented environment increases administrative burdens, delays service delivery, and undermines the integrity of national educational statistics.

To address these systemic challenges, Tanzania has initiated the deployment of Jamii X-Change, a national digital interoperability platform designed to enable secure, real-time data exchange across public institutions. Jamii X-Change facilitates efficient communication between schools, examination councils, and civil registration authorities by standardising data formats and authentication protocols. Within the education sector, the platform plays a critical role in automating student identity verification, supporting accurate candidate registration, and enhancing the reliability of data used for planning and policy formulation [4][6]. The operational structure and functionality of the platform, illustrated in Figure 1, demonstrate how Jamii X-Change functions as a central hub linking student information systems, national identification databases, and examination management systems into a unified and secure data flow. Through these integrations, Jamii X-Change not only streamlines assessment processes but also establishes the foundation for a more integrated, responsive, and data-driven educational infrastructure.

Figure 1. Jamii X-Change for education assessment.

Building on this foundation, this study proposes the design of a student data lake architecture model that leverages the Jamii X-Change platform. A student data lake functions as a centralised and scalable repository capable of storing large volumes of raw, structured, and unstructured data from diverse sources, including school records, national examination results, attendance logs, and demographic information. Unlike traditional relational databases that enforce fixed schemas at the point of data ingestion, data lakes support a schema-on-read approach, thereby enabling flexible analytics and real-time reporting [7].

The proposed architecture has the potential to strengthen Tanzania’s educational assessment ecosystem by enabling accurate and timely access to student records, reducing data duplication, streamlining examination processes, and supporting dynamic reporting and evidence-based policy interventions. Nevertheless, the implementation of such an architecture presents several challenges, including the need for reliable digital infrastructure, robust data governance frameworks, and adequate technical capacity among stakeholders to manage and interpret data effectively. Moreover, successful deployment requires alignment with national education policies, as well as sustained collaboration among government agencies, educational institutions, technology providers, and civil society organisations. These considerations underscore the importance of a holistic, multi-stakeholder approach to ensure the sustainable adoption and long-term impact of the proposed system.

ResearchObjective

The main objective of this study is to design and evaluate a student data lake architecture model leveraging the Jamii X-Change platform, with the aim of enhancing the efficiency, security, and inclusivity of educational assessment processes in Tanzania.

ResearchQuestions

1) How can a student data lake architecture model, built on Jamii X-Change, improve the accuracy, efficiency, and security of educational assessment in Tanzania?

2) What are the key architectural components and integration strategies necessary for implementing a scalable and interoperable student data lake within Tanzania’s education sector?

3) What infrastructural, technical, and governance challenges may arise in the deployment of a student data lake using Jamii X-Change, and how can these be mitigated?

4) How do stakeholders-including educators, administrators, policymakers, and technical staff-perceive the benefits and limitations of a student data lake architecture model for educational assessment?

This research study addresses the challenges posed by fragmented, limited-analytics, and non-interoperable educational assessment systems in Tanzania. The study systematically reviews relevant literature, identifies key research gaps, and outlines the research methodology, with findings and discussions organized around the study’s research questions. It further proposes a scalable and secure Student Data Lake architecture designed to improve the accuracy, efficiency, and reliability of educational assessment processes. Moreover, the study examines the policy implications of implementing such a system in Tanzania and provides practical recommendations and concluding insights to support the development of a resilient, interoperable, and evidence-based digital assessment ecosystem.

Student data lakes have increasingly become essential as educational institutions seek scalable and flexible architectures for integrating and analysing large volumes of data from multiple, often fragmented, sources. A data lake serves as a centralised repository that stores both raw and processed data in diverse formats, thereby enabling advanced analytics, real-time reporting, and machine learning applications [8]. In the education sector, such architectures facilitate the consolidation of student records, assessment results, attendance information, and demographic data, empowering institutions and policymakers with timely and accurate insights to improve learning outcomes and optimise resource allocation [9]. The inherent flexibility of data lakes makes them more suitable than traditional data warehouses for national education systems characterised by heterogeneous data types and evolving reporting requirements.

Several institutions have demonstrated the effectiveness of student data lake implementations. For instance, Ohio State University’s Reporting and Analytics Environment integrates diverse institutional datasets into a governed analytics platform accessible to faculty members, administrators, and researchers [10]. Similarly, the New York City Department of Education employs a cloud-based data lake to integrate real-time school data, thereby enabling district-level monitoring and timely interventions [11]. These implementations illustrate how centralised and interoperable data environments enhance transparency and support evidence-based decision-making in education. A critical enabler of such data lake functionality is the underlying interoperability framework, which ensures secure, standardised, and real-time data exchange between heterogeneous systems.

Leveraging Jamii X-Change presents a strategic opportunity to consolidate fragmented education data sources, including student registration systems, examination boards, and teacher databases, into a unified repository. This integration supports longitudinal analysis, timely reporting, and improved credential verification across Tanzania’s education ecosystem. Importantly, it addresses persistent challenges within the country’s education system, such as manual data processing, limited system interoperability, and data accuracy issues that continue to constrain effective policy formulation and service delivery.

Emerging empirical research supports the development of student data lakes tailored to the Tanzanian context. Existing studies emphasise the need for scalable and context-sensitive data architectures capable of accommodating diverse educational environments characterised by variable digital infrastructure and pronounced rural-urban disparities [12]. Data lake architectures incorporating modular components, such as data ingestion pipelines, anonymisation modules, and analytics dashboards, enable phased implementation and functional flexibility, making them particularly suitable for resource-constrained environments [13]. Furthermore, the integration of Jamii X-Change Metrics enables real-time monitoring of data exchange health and service performance, which is essential for maintaining stakeholder trust and ensuring long-term system reliability.

Despite these insights, a notable gap remains in the academic literature regarding the evaluation of student data lakes built on Jamii X-Change for improving educational assessment processes in Tanzania. Existing ICT research in the country has largely focused on infrastructure limitations, digital literacy, and isolated automation initiatives, without addressing comprehensive and interoperable data ecosystems that span the full assessment lifecycle. Moreover, the potential of Jamii X-Change to facilitate secure, real-time data interoperability has not been sufficiently leveraged or empirically examined within the context of educational data lakes. International best practices underscore the importance of robust data governance frameworks, security protocols, and sustained human capacity building in supporting successful digital transformation initiatives [3][4]. Applying these principles to the design of a student data lake architecture underpinned by Jamii X-Change can help overcome existing challenges, reduce manual interventions, and provide policymakers and educators with reliable, actionable data. Accordingly, this study seeks to address this research gap by developing and evaluating a scalable and secure student data lake model that leverages Jamii X-Change’s interoperability features and metrics capabilities to enhance educational assessment and data-driven decision-making in Tanzania.

Based on the results findings and stakeholder feedback, a student data lake architecture is proposed (Figure 5) to further address the challenges of data fragmentation and limited analytics capacity. The system design leverages Jamii X-Change for secure, interoperable data exchange and incorporates X-Road Metrics for operational monitoring. The main modules are:

1) DataIngestionModule: Responsible to retrieve student data from Security Servers of NECTA, RITA, TCU, VETA, Loans Board, and other sources via Jamii X-change. The student data retrieved shall be inserted to the Mongo database. The execution of the collector module is performed automatically via a cron job task. The collector module needs HTTP-access to the Jamii X-change Central Server to get from global configuration list of members Security Servers. The collector module needs HTTP-access to an X-Road Security Server that acts as an Operational Monitoring Client to get the data collected. The collector module needs access to the Mongo Database.

2) Datacorrectormodule: The Corrector module is responsible to clean the student data and derive monitoring metrics in a clean database collection. The execution of the corrector module is performed automatically via a service task. It is important to note that it can take up to 7 days for the Collector module to receive operational data from (all available) Security Server(s) and up to 3 days for the Corrector module to clean the students raw data and derive monitoring metrics in a clean database collection.

3) Anonymizer: Is responsible of preparing the operational monitoring data for publication through the Opendata module. Anonymizer configuration allows Jamii x-change Metrics extension administrator to set fine-grained rules for excluding whole operational monitoring data records or to modify selected data fields before the data is published. The anonymizer module uses the operational monitoring data that Corrector module has prepared and stored to MongoDb as input. The anonymizer processes the student data using the configured rule-set and stores the output to the open data PostgreSQL database for publication (Figure 3).

Figure 3. Anonymizer module.

MongoDb is used to store “non-anonymized” operational monitoring student data that should be accessible only by the Jamii X-Change Metrics administrators. Anonymized operational monitoring students’ data that can be published for different stakeholders is stored in the PostgreSQL. The Opendata UI needs access only to the PostgreSQL (Figure 3). To follow the “principal of least privilege” it is recommended to install Opendata UI on a dedicated host that has no access at all to MongoDb. However, the Anonymizer module needs access also to the “not-public” data, so it should run on a host that has access to both MongoDb and PostgreSQL.

4) Networking/VisualizerModule: The purpose of the Networking module is to visualize the networking activity between the Jamii X-Change members (Figure 4). It consists of:

Data preparation R script that queries the Open data PostgreSQL database, does relevant calculations (query counts between Jamii X-Change members) and writes a table file to be used in the visualization web application.

Also networking module consists of RStudio Shiny-based web application to visualize the networking activity between X-change members in education sector.

Periodic crony job run to execute data preparation script. This can be started with the following steps: Read configuration parameters from a settings file, Read instance specific complementary files; establish a connection to Open data

Figure 4. The general scheme of processes in the Networking module.

PostgreSQL and queries the most recent date in field; querie the most recent data from Open data dating back to certain number of days (“interval” in settings file, default 30 days) from the most recent data. The script retrieves only succeeded = True logs. The script retrieves the following fields: Calculates the count of service calls between each unique combination of the fields: Add the following concatenate fields client, producer, and producer. Service using newline escape sequence as a separator: Parses member names from riha.json and adds the names to the output data, writes the resulting table to RDS-file (R-specific binary file) in the shiny application’s dynamic data directory.

5) Reportsmodule: Is responsible for creating monthly reports about subsystems of X-Change members in education sector (datasets usage). The execution of the report’s module can be either performed automatically (via cron job) or manually. Overall system, its users and rights, processes and directories are designed in a way, that all modules can reside in one server (different users but in same group “Jamii X-change-metrics”) but also in separate servers. Also, the System designed in a way, that allows to monitor data from different X-change instances. However, the system designed in such a way that it can be used by all X-change members in education sectors as well as for Member own monitoring (includes possibilities to monitor also members data exchange partners). The proposed System architecture design for a Student Data Lake integrated with Jamii X-Change is presented in Figure 5.

This architecture (Figure 5) is modular, scalable, and designed to support both current and future data analytics needs in Tanzania’s education sector. The proposed design for a Student Data Lake integrated with Jamii X-Change provides a practical solution for managing and gradually phasing out outdated RITA

Figure 5. System architecture design for a Student Data Lake integrated with Jamii X-Change.

typewritten birth certificates. It converts them into authoritative, digitally verifiable identity records. Legacy birth certificate data can be digitised and ingested into the data lake through controlled capture processes. These processes standardise key attributes such as name, date of birth, registration number, place of birth, and issuing authority, and link them to student identifiers. Once ingested, the system can securely validate and reconcile these records against e-RITA registration and national identification systems. This reduces identity mismatches, duplicate records, and delays associated with manual verification during examination registration and results processing.

The architecture supports a phased deprecation approach, whereby legacy certificates remain accessible for audit purposes while new registrations rely on digitally authenticated identity data. This enables RITA to progressively replace typewritten certificates with trusted, interoperable digital records, thereby enhancing accuracy, efficiency, and integrity across Tanzania’s educational assessment system. The approach minimises disruption to existing workflows, allowing institutions to maintain continuity during the transition. It also establishes a foundation for future integration with national digital identity systems, ensuring long-term scalability and compliance.

Furthermore, as data from additional institutions such as TCU, NACTVET, VETA, TIE, HESLB, MoEST, PMO‑RALG, and higher learning institutions in Tanzania are integrated, a robust governance model ensures that data ownership remains with the originating institutions. At the same time, access rights and usage permissions are managed centrally through role‑based access controls and comprehensive audit logging, which reinforces accountability, security, and compliance with national data policies, while supporting both day‑to‑day operational requirements and the long‑term sustainability of the system.

From a policy perspective, the adoption of a Jamii X-Change-enabled student data lake has significant implications for educational assessment in Tanzania. First, it provides an evidence-based foundation for targeted resource allocation, enabling the Ministry of Education to prioritise investments in ICT infrastructure for underserved regions, particularly rural and hard-to-reach schools where data gaps are most pronounced. Real-time, integrated data on enrolment, assessment outcomes, and teacher deployment can support more equitable distribution of learning materials, examination resources, and qualified educators, consistent with national goals for inclusive and equitable quality education [3][21].

Second, the successful operationalisation of such a data lake necessitates deliberate policy commitment to digital literacy and capacity building, including continuous training for education officers, school administrators, and ICT personnel in data management, analytics, and information security. Empirical evidence suggests that investments in digital skills are critical for translating ICT infrastructure into improved institutional performance and data-driven decision-making [4][22]. Without parallel investments in human capital, the benefits of advanced data architectures may remain underutilised. Finally, embedding the student data lake within national education policy frameworks would strengthen data governance, interoperability standards, and accountability mechanisms, ensuring sustainability and alignment with Tanzania’s broader digital government and education transformation agendas, including e-Government and cross-sector data sharing initiatives [3].

1) PhasedNationwideRollout: Expand Jamii X-Change and the student data lake through staged pilots, starting in regions with varying infrastructure and scaling up based on lessons learned.

2) CapacityBuilding: Implement ongoing digital literacy and technical training for educators, administrators, and IT staff.

3) InfrastructureInvestment: Prioritize internet connectivity and device access, especially in rural and underserved areas.

4) RobustDataGovernance: Develop and enforce clear policies on data privacy, security, and access control.

5) ContinuousMonitoring: Use X-Road Metrics and regular audits to ensure system health, guide improvements, and maintain stakeholder trust.

6) StakeholderEngagement: Involve all relevant parties in design, implementation, and evaluation to ensure buy-in and address context-specific needs.

This study demonstrates that Jamii X-Change technology, when integrated with a well-designed student data lake, can significantly enhance the efficiency, accuracy, and accessibility of educational assessment data in Tanzania. The pilot results and stakeholder feedback confirm that interoperability, automation, and centralized analytics are key to overcoming persistent challenges. The proposed architecture offers a scalable, secure solution that aligns with international best practices and supports evidence-based decision-making. For successful nationwide adoption, sustained investment in infrastructure, comprehensive training, and phased implementation are essential. Future work should focus on expanding the data lake to include additional educational and social data sources, further enhancing analytics for policy and planning.