Why Turkey’s Election Infrastructure Runs on Oracle: A Deep Dive into SEÇSİS and Database Choices

When Professor Özgür Demirtaş posted a short but striking message on X, questioning what Oracle actually does and why it continues to thrive despite appearing distant from everyday tech discourse, it hit a nerve. To many, Oracle is an enigma—a giant of the enterprise world, yet invisible in our daily digital interactions. It doesn’t make phones or shiny apps. It doesn’t trend. But it's everywhere power and data converge. His remark made me pause and reflect: what is it that Oracle really powers? And in Turkey, the answer turns out to be quite specific. Oracle helps run SEÇSİS, Turkey's national election information system.

This realization led to a deeper look into how elections are operationalized in Turkey. SEÇSİS is the centralized system that handles the administrative and reporting side of elections. It doesn’t count votes electronically, nor does it replace physical ballots. Instead, it processes and aggregates election-related data such as voter registration, candidate lists, and scanned ballot box results. It acts after the manual counting process, when signed documents and figures are sent from local election boards to central systems.

The infrastructure is layered. Local election officials input data manually into district-level computers. That data is transmitted through dedicated networks to central servers in Ankara. These servers store and process the information using a relational database system. The design is centralized but distributed in responsibility—local actions feed a national outcome.

The database behind this infrastructure is Oracle. The choice of database technology appears to rest on a few consistent needs:

• High availability, especially during nationwide elections when millions of inputs may occur over a short period.

• Transaction integrity, so that each result entry is atomic and fully recorded.

• Centralized support and accountability, especially under public scrutiny.

According to publicly available reports, SEÇSİS uses Oracle’s RAC (Real Application Clusters) to run its database. This allows for multiple servers to coordinate access to a single dataset, ensuring fault tolerance and distributed load. SEÇSİS’ central databases are reportedly isolated from the internet and accessible only through a secure network that is managed in conjunction with other government infrastructure.

At the application level, the software interacting with the database appears to be developed in Java EE, running on Linux servers. Client terminals in local election boards, often located in courthouses, operate on Windows. The overall architecture separates data entry environments from public-facing results portals. There are read-only database replicas used for displaying election data to political parties and the public.

Key operational principles include:

• Manual data entry, based on physical documents like signed ballot box results

• Document scanning, to retain digital copies of wet-signed records

• One-way replication, from secure internal systems to external portals

• Role-based access, limiting who can interact with which parts of the system

What’s notable about SEÇSİS is not a reliance on new or experimental technology, but its tight integration into the bureaucratic workflow. Every input is cross-referenced. Signed documents are scanned and stored. Election results, once entered, are linked to their original physical form. These steps are traceable through logs, snapshots, and user access trails.

The system is not without critique. There are concerns about oversight, transparency, and access controls. Technical secrecy around the system’s configuration and the lack of independent audits have raised questions in public discourse. These are political and institutional questions as much as technical ones. Yet from a technology perspective, the choices appear rooted in long-term viability and control rather than in agility or openness.

The database system—in this case Oracle—was likely chosen for these reasons:

• It is already embedded in existing public IT infrastructure.

• It supports clustering and availability without requiring third-party tools.

• It comes with vendor guarantees and certification mechanisms.

Alternatives such as PostgreSQL or SQL Server, while viable, would require different assumptions about support, certification, and systemic rewrites. Migration from Oracle would mean rewriting not just software, but governance procedures, documentation, training material, and internal workflows.

None of this is an endorsement. It's a technical description. It outlines the weight of institutional momentum and the infrastructure decisions that are difficult to reverse without systemic change. SEÇSİS, as currently implemented, reflects a centralized and highly managed model of information processing. Oracle is a part of that model—not necessarily because it is the only choice, but because it has been embedded over time.

There are broader implications. As elections become more complex and digitally recorded, the design of systems like SEÇSİS will continue to shape the transparency, security, and public perception of democratic processes. What happens inside that database matters—not because of who makes it, but because of what it stores and how that information flows through layers of law, infrastructure, and social trust.