Public Sector
Cities today are under immense pressure—facing tighter budgets and increasing demands for public services, making effective community engagement more essential than ever. Onetera, a leader in civic technology, set out to help municipalities serve residents with greater speed, accuracy, and intelligence. By leveraging NVIDIA RTX PRO™ 6000 Blackwell Workstation Edition GPUs and NVIDIA TensorRT™-LLM optimization, Onetera’s Government Intelligence Agent, which provides AI-powered support for accessing critical government services, streamlines complex processes, accelerates inference, and enables rapid deployment of critical programs at scale. Onetera leverages local inference primarily for preprocessing of unstructured public data. By preprocessing scattered municipal data (across PDFs, city websites, county databases, etc.), Onetera can more efficiently and reliably serve conversational AI experiences at runtime.
Onetera
Accelerated Computing Tools & Techniques
Local governments sit at the forefront of community support, fielding everything from housing crises and business permits to emergency response and economic recovery. Yet outdated technology stacks and traditional systems often fail to keep pace with rapidly shifting circumstances, leaving city staff to navigate overwhelming volumes of regulation, data, and resident requests all at once.
Onetera recognized that fragmented municipal data, slow manual workflows, and the absence of scalable solutions all raised costs and reduced staff capacity to support those in need.
Faced with the unique challenge of powering AI-driven municipal programs for more than 90 cities, Onetera set out to create a platform that could instantly ingest, process, and deliver actionable intelligence across tens of thousands of pages of diverse municipal codes, zoning maps, permit requirements, and business rules simultaneously—without requiring heavy IT overhead. Legacy compute resources and traditional IT approaches couldn't keep up, threatening Onetera's promise of "zero IT lift" and rapid deployment for city partners. Onetera's local inference server uses a variety of open-source models for different tasks, such as GPT OSS 20B and Llama 4, providing the flexibility and control needed to tailor solutions to each municipality's specific requirements.
By leveraging the massive parallel processing and 96 GB memory capacity of NVIDIA RTX PRO 6000 Blackwell Workstation GPUs, Onetera accelerates the analysis of zoning maps, permit logs, and regulatory records, batch processing thousands of complex documents across jurisdictions in real time, all within a secure on-premises environment.
Onetera's breakthrough came with the development of its Government Intelligence Agent—a central AI-powered foundation capable of delivering sustained innovation at municipal scale—powered by NVIDIA RTX PRO 6000 Blackwell Workstation Edition GPU and TensorRT-LLM acceleration.
Leveraging next-generation computing capabilities, this agent can process complex, multi-format municipal datasets from dozens of cities simultaneously, extracting structured intelligence that seamlessly powers a growing suite of city programs.
By integrating advanced hardware acceleration with the NVIDIA RTX PRO 6000 Blackwell Workstation Edition GPU and optimized inference via TensorRT-LLM, Onetera's ability to serve municipalities transformed by enabling:
The incredibly high throughput of RTX PRO 6000 GPUs allows Onetera to support dozens of municipalities simultaneously, while TensorRT-LLM ensures that large-scale language model inference is affordable, reliable, and optimized for their municipal workloads.
This robust AI foundation not only powers high-demand programs-like business license guides, permit assistance, and rental registries-it also provides the agility to launch new solutions as municipal needs evolve. The success of initiatives like the Permit Service Guide, first launched for Altadena residents rebuilding after the Los Angeles wildfires, showcases how quickly Onetera's approach delivers tangible benefits to communities in crisis.
"NVIDIA RTX PRO 6000 Blackwell Workstation Edition GPUs enable our Government Intelligence Agent to be a true force multiplier for municipal innovation," said Felix Ruano, Onetera CEO. "We're able to gather and process public information at a scale and speed not previously possible, giving us a powerful foundation that can simultaneously serve business guides, permit assistance, housing programs, and whatever cities need next. This is exactly the kind of AI-native process that creates large-scale transformative community impact."
Ontera
With its Government Intelligence Agent at the core, Onetera has fundamentally reshaped how cities deliver services. Programs that once took months to launch can now be deployed in days, allowing cities to respond dynamically to evolving resident needs and emergencies.
By eliminating redundant processes and leveraging real-time data accuracy, cities like Escondido can better streamline operations and ensure every recommendation reflects the latest regulations.
In Escondido, one of the biggest hurdles for economic development has been the complexity of zoning regulations. According to Jennifer Schoeneck, Director of Economic Development for the City of Escondido, businesses often commit to properties without fully understanding the requirements, only to face costly delays, conditional use permits, or even outright denials.
For example, a screen printing shop that promised new jobs and sales tax revenue was prevented from opening downtown because the specific plan didn't explicitly list that use-even though it could reasonably fall within existing categories. The problem wasn't a lack of information; it was that critical details were buried deep within dense PDFs, inaccessible to brokers and business owners who needed them to make decisions.
With Onetera's AI Agent integrated into the Business Service Guide, Escondido staff can now surface this information quickly and digitally, giving brokers and tenants a clear view of what's possible before committing to a space. For the Business Service Guide and related government use cases, Onetera uses LangGraph as its framework to define agent flow and tool-calling. For each agent, the team defines the intended goals and conditional logic across nodes, then ensures the agent has access to the correct quality-controlled, preprocessed data through context engineering. For more complex workflows, such as the ADU Permit Agent, Onetera chains together multiple capabilities (including spatial AI analysis) while still ensuring a target path using LangGraph. This approach provides users with a conversational, flexible experience while ensuring accuracy and reliability of information.
This prevents costly missteps while helping businesses make informed decisions faster and avoid unnecessary setbacks. “If we could give brokers and tenants this information digitally before they commit to a space, we'd save everyone time and prevent those frustrating situations where dreams meet regulatory reality,” Schoeneck said.
"NVIDIA RTX PRO 6000 Blackwell Workstation Edition GPUs enable our Government Intelligence Agent to be a true force multiplier for municipal innovation."
Felix Ruano
Onetera CEO
For Onetera and its city partners, the business value is clear: scalable innovation, powerful efficiency, and adaptability to address challenges as they arise-without overwhelming technical, operational, or financial barriers.
Onetera and NVIDIA demonstrate that digital transformation in government is not just possible—it's affordable, scalable, and already delivering real results for cities and residents. By fusing RTX PRO 6000 Blackwell Workstation Edition performance with TensorRT-LLM inference optimization, Onetera is making AI-native city service delivery a practical, powerful force for public good.
Learn more about NVIDIA RTX PRO GPUs and how they can accelerate your workflows.