Customer Service Automation · Document Processing
Onetera needed to help more than 90 municipalities turn fragmented regulations, zoning maps, permit rules, and legacy documents into fast resident-facing services without imposing heavy IT overhead or expensive manual processing on city partners.
Reported impact
Onetera reduced end-to-end municipal program processing time by about 40 percent, cut per-run costs by 88 percent, increased throughput 1.7x, supported dozens of municipalities simultaneously, and shortened program launches from months to days for services such as permits and business guides.
Onetera built its Government Intelligence Agent as a local-inference stack for municipal data using NVIDIA RTX PRO 6000 Blackwell Workstation Edition GPUs and NVIDIA TensorRT-LLM. The platform preprocesses unstructured public data from PDFs, city websites, county databases, and other records on premises so it can serve reliable conversational experiences at runtime. Onetera combines open models such as GPT OSS 20B and Llama 4 with LangGraph-driven agent flows, where each agent has defined goals, tool-calling paths, and access to quality-controlled context. For more complex workflows, including permit agents and spatial-AI use cases, the system chains multiple capabilities while keeping humans in the loop for validation. This architecture gives Onetera a reusable municipal AI foundation that can launch new city programs quickly while controlling cost and maintaining legal accuracy.
Municipal data from PDFs, websites, county databases, zoning maps, and permit logs is preprocessed locally on RTX PRO systems, LangGraph routes tasks through the Government Intelligence Agent and open models such as GPT OSS 20B or Llama 4, and the resulting structured guidance is returned through resident and staff service experiences.
By preprocessing scattered municipal data (across PDFs, city websites, county databases, etc.), Onetera can more efficiently and reliably serve conversational AI experiences at runtime.
For the Business Service Guide and related government use cases, Onetera uses LangGraph as its framework to define agent flow and tool-calling.
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's local inference server uses a variety of open-source models for different tasks, such as GPT OSS 20B and Llama 4.
Programs that once took months to launch can now be deployed in days.
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
NVIDIA Sample Jun 26, 2026 MISA MISA needed to move from expensive, externally hosted LLM APIs to a self-hosted Vietnamese AI stack that could protect sensitive data, reduce latency and operating cost, and support many concurrent public-service and… $200,000 in monthly cost savings MISA AVA · NVIDIA A40 GPUs · NVIDIA DGX H200 
OpenAI Sample Jun 12, 2026 Preply Preply wanted to personalize language learning more deeply and reduce repetitive tutor work without replacing the human relationship that makes one-to-one teaching effective. 4.7/5 satisfaction rating Brand Voice GPT · ChatGPT Enterprise · Codex 
Google Cloud Sample Archived · Feb 11, 2026 Wonderful Wonderful was looking for a technology partner that could combine advanced multilingual models with globally distributed infrastructure, at a cost that scaled with their growth. They also needed strong reliability,… Gemini powers expansion into 30 multilingual markets Cloud Storage · Gemini · Google Kubernetes Engine (GKE) 