SF's Duplicate Image Problem: The Key Decisions That Will Shape the City's Digital Infrastructure

San Francisco's municipal technology office has identified a sprawling backlog of duplicate image files spread across department servers citywide, a years-in-the-making storage problem that officials say is now actively hampering everything from permit processing at the Department of Building Inspection on Freemont Street to case management at the Department of Public Health's Zuckerberg San Francisco General Hospital campus. The question now is what happens next — and who decides.

The timing matters. The city is mid-cycle on a broader digital infrastructure overhaul that began under Mayor London Breed's administration and has carried into the current fiscal year. San Francisco's Department of Technology, headquartered at 1 South Van Ness Avenue, has until September 30 to submit its consolidated storage roadmap to the Budget and Legislative Analyst's office before Q1 2027 capital allocations are locked in. Miss that window and the duplicate-image cleanup loses its dedicated funding line entirely.

What the Backlog Actually Looks Like

The problem is more mundane than it sounds, and more expensive. Duplicate image files — scanned permits, case photos, inspection records, health documents — accumulate when departments run parallel workflows without shared asset-management systems. Each copy costs money to store, back up, and migrate. City IT staff who reviewed internal storage audits earlier this year found some departmental folders contained three or more identical versions of the same file, a pattern consistent across at least a dozen agencies. The San Francisco Public Library's digital archive division at the Civic Center branch flagged a similar issue in its digitized historical collection as far back as 2023, with no resolution since.

Municipal cloud storage isn't cheap. Enterprise-tier contracts with major providers typically run between $0.02 and $0.08 per gigabyte per month depending on redundancy requirements — costs that multiply fast when tens of thousands of duplicate files sit untouched across a system the size of San Francisco's. The Department of Technology manages infrastructure for roughly 60 city departments and commissions.

The SF Digital Services team, which operates separately from the main IT department and focuses on resident-facing platforms, has already piloted an automated deduplication tool on a subset of Planning Department records stored on servers supporting the 49 South Van Ness permit portal. Early results from that pilot, which ran through May 2026, showed meaningful reduction in redundant files within the tested file set — but scaling that tool citywide requires both procurement approval and coordination across agencies that historically guard their own data systems.

The Decisions Ahead — and Who Makes Them

Three choices will define the outcome between now and year's end. First, the Department of Technology must decide whether to standardize on a single deduplication vendor or allow departments to procure their own tools independently. Independent procurement is faster but almost guarantees incompatibility down the line. Second, the city's chief data officer — a role that has seen significant turnover since 2022 — needs to issue clear data-governance rules on image retention periods. Without those rules, deduplication tools simply flag files without authority to delete anything. Third, the Board of Supervisors' Government Audit and Oversight Committee, which last convened on digital infrastructure matters in March 2026 at City Hall's Room 250, must decide whether to hold hearings that give the cleanup political momentum or let it languish as a bureaucratic backroom exercise.

The SF Digital Services team is expected to present preliminary options to department heads by late July. If the standardized-vendor path is chosen, a request for proposals would need to go out no later than August to meet the September budget deadline. If departments are allowed to go their own way, the timeline loosens — but so does any hope of a coherent, citywide result.

For residents, the practical stakes are modest but real. Faster permit lookups, more reliable case records at DPH clinics, and a digital archive at the Main Library that actually functions without redundant clutter are the tangible downstream benefits. San Francisco has been here before: the city launched a data-consolidation initiative in 2019 that stalled within 18 months. This time, the budget clock is already running.