NYC Startup Raises $45 Million to Rebuild City's 311 System With AI

NEW YORK — A Manhattan-based government technology startup called CivicAI announced this week that it has raised $45 million in a Series B funding round led by venture capital firm Andreessen Horowitz, with participation from two New York City pension funds and the NYC Economic Development Corporation's venture arm. The company is developing an AI-powered replacement for the city's 311 non-emergency services system — a platform that handles more than 20 million calls and digital contacts per year but has been widely criticized by residents and city employees alike for slow resolution times and inconsistent routing.

CivicAI's platform replaces the existing 311 call center workflow with a conversational AI assistant capable of understanding and responding to service requests in 14 languages, automatically routing complaints to the correct city agency and sub-unit, and generating follow-up notifications to residents when their requests are processed. The company says its routing accuracy in a pilot conducted with the Department of Sanitation and the Department of Buildings last year reached 94 percent on the first attempt, compared with 71 percent for the existing 311 system — a gap that translates directly into resolution speed.

The $45 million will fund a full-scale rollout across six city agencies beginning in the fall, the company said, as well as the expansion of a predictive analytics module that identifies geographic clusters of recurring service complaints before they escalate. That module, which the company calls the Signal Layer, synthesizes 311 data, 911 call logs, infrastructure inspection records, and social media noise to generate early warning flags for agency supervisors. In the pilot, the Signal Layer identified three areas of elevated pothole density and two clusters of rodent activity complaints seven to ten days before those areas generated peak 311 volume.

"311 is where residents experience city government most directly," said Amara Diallo, CivicAI's founder and CEO, who previously worked as a deputy commissioner in the Mayor's Office of Operations. "When the system works — when your noise complaint is routed correctly and you get a callback — people trust that the city is responsive. When it doesn't, that trust erodes. We built this to close that gap at scale." Diallo said the company has been in conversations with three other major American cities about adapting the platform to their non-emergency services infrastructure.

The city's existing 311 system, which was built in 2003 and has undergone incremental updates, handles approximately 22 million annual contacts across phone, web, and mobile channels. Staffing the call centers costs the city roughly $180 million per year, a figure that CivicAI says can be reduced by 35 to 40 percent through AI-assisted triage while maintaining or improving response quality. The company is careful to frame the platform as an augmentation of human staff rather than a replacement, noting that complex, sensitive, or escalated cases would continue to be handled by trained call center agents.

Data privacy advocates and city employee unions have both raised questions about the rollout. The technology arm of District Council 37, which represents many of the city employees who staff 311, has requested assurances that the platform will not be used to reduce headcount, and has asked for independent audits of how the conversational AI handles sensitive complaints, including housing emergencies and reports of domestic situations. CivicAI said it will work with union representatives and the city's Department of Citywide Administrative Services on implementation agreements before the fall rollout begins.

The broader trend of AI deployment in city government is accelerating nationally, with municipalities from Los Angeles to Boston experimenting with AI tools for permit processing, pothole detection, and benefits eligibility screening. New York, with its scale and the density of its resident-government interaction, is in many ways the highest-stakes test case. Diallo said the Series B will also fund a governance and accountability team dedicated to auditing the platform's outputs for bias and accuracy — a function she described as non-negotiable for a product that mediates between residents and public services.