São Paulo's Duplicate Image Problem: The Key Decisions Ahead for the City's Digital Archives

São Paulo's municipal government is sitting on a sprawling mess of duplicated digital images across at least three major public databases, and the window for a clean, cost-effective fix is narrowing fast. The Secretaria Municipal de Gestão (SMG), which oversees the city's information infrastructure, has been wrestling internally since early 2026 with how to handle tens of millions of redundant image files clogging systems used by departments ranging from urban planning to social services. The problem is not new, but pressure to resolve it has intensified as the city moves to consolidate its digital services under the São Paulo Inteligente program ahead of a planned 2027 platform migration.

Why now? The city has set a hard internal deadline of the first quarter of 2027 for the full integration of legacy systems into a unified cloud framework managed through a contract with state-owned infrastructure partner PRODESP. Every duplicate image file that survives into that migration costs more to store, index, and secure. Storage costs for municipal cloud infrastructure have climbed alongside global data centre pricing — and São Paulo's own budget for digital transformation, allocated at R$480 million across the 2025-2026 biennium under the city's Programa de Modernização da Gestão, leaves limited room for bloated data overhead.

Where the Bottlenecks Are Forming

The most acute concentration of the problem sits in two places: the GeoSampa geospatial platform, which holds aerial and street-level imagery of every neighbourhood from Pinheiros to Cidade Tiradentes, and the SIURB system used by the Secretaria de Infraestrutura Urbana to document construction permits and drainage works across the city's 96 subprefectures. Engineers working on both platforms have identified redundancy rates that make deduplication a prerequisite for any serious performance improvement, though the exact scale of the problem has not been publicly disclosed by the city.

Private sector players are already circling. At least two São Paulo-based tech companies with roots in the Cubo Itaú startup hub in Faria Lima have pitched automated deduplication tools to the SMG this year. The pitch is straightforward: machine-learning models trained on municipal image metadata can flag duplicates faster and more cheaply than manual audits. The decision before Mayor Ricardo Nunes's administration is whether to procure that technology through a standard licitação process — which can take six months or more — or to fast-track a pilot under the city's existing Marco Legal das Startups provisions, which allow for lighter-touch contracting with emerging tech firms.

Three Decisions That Will Shape the Outcome

The first and most immediate question is governance: who inside the city structure owns the deduplication mandate? Right now, GeoSampa sits under the Secretaria Municipal de Urbanismo e Licenciamento, while SIURB reports to a separate secretariat entirely. Without a single coordinating authority, past interoperability projects in São Paulo — including a 2022 attempt to unify permit records across subprefectures in the Zona Leste — stalled precisely because no agency wanted to absorb the liability for data loss during transitions.

The second question is legal. Municipal image archives include photographs tied to enforcement actions and social program eligibility checks. Deleting what appears to be a duplicate without airtight provenance tracking could violate Brazil's Lei Geral de Proteção de Dados Pessoais, the LGPD, which came into force in August 2020 and imposes strict requirements on the handling and destruction of personal data. The Autoridade Nacional de Proteção de Dados in Brasília has already issued guidance to state and municipal bodies about data minimisation obligations, and São Paulo's legal department will need to sign off on any deletion protocol before engineers touch a single file.

The third question is money. Even a well-scoped deduplication project requires staff time, legal review, vendor contracts, and audit trails. City hall's technology budget for the second half of 2026 is already committed largely to the PRODESP migration preparatory work and to a cybersecurity upgrade for the Poupatempo digital portal on Avenida Paulista. Any serious deduplication initiative will require either a budget supplement or a reallocation — a political decision that sits with the Câmara Municipal.

The coming weeks matter. If the SMG does not designate a lead agency and begin a formal procurement process before September, the PRODESP migration timetable becomes functionally impossible to meet without carrying the duplicate data forward — and paying to store it for another budget cycle. City technology officials have been meeting with counterparts from the Governo do Estado since June to map out the migration sequence. The deduplication question is on the agenda. How prominently it features in those conversations will determine whether São Paulo enters 2027 with cleaner, faster public systems or simply a bigger, more expensive version of the same problem.