NERHIA: The Urban Nervous System

NERHIA: The Urban Nervous System is an AI-native operational service for small physical-digital businesses and urban-service operators. These organizations work with signals from equipment, public sources, customer reports, field observations, and internal systems, but the information is usually fragmented. A small operator may know that something changed without having the time or specialist staff to determine why it matters, which evidence is trustworthy, what information is missing, and what should happen next. NERHIA converts that fragmented evidence into a structured, auditable command brief that a human operator can review.

The project competes in Small Business Services because its purpose is to give smaller operators access to an operational capability normally reserved for organizations with dedicated data, AI, and control-room teams. A user provides or selects a scoped signal set. NERHIA normalizes source, time, territory or asset, confidence, known uncertainty, related evidence, permitted actions, and restrictions. It then calls Gemini through VBC Compute Layer. Gemini interprets that bounded evidence package and produces a briefing that explains what changed, why it may matter, which conclusions are observations versus inferences, what remains uncertain, and the next verification step.

AI is part of the operating workflow rather than an ornamental chat feature. It performs evidence interpretation and briefing work that would otherwise require a specialist to reconcile multiple inputs manually. The protected production path sends an explicit Gemini request through VBC on Google Cloud Run with paid-provider fallback disabled and cache bypassed. VBC records provider and model metadata, enforces idempotency, checks that repeated delivery does not charge twice, and produces a redacted receipt. The receipt is designed to prove execution without exposing prompts, model output, credentials, raw provider errors, or raw user identifiers. Firebase Hosting provides the public NERHIA experience at nerhia.com.

The business is designed to create work beyond the founding role rather than removing human authority. Each deployment can require local sensor installers, data-integration technicians, field inspectors, operator analysts, customer-success support, and channel partners who understand the physical site. Gemini compresses evidence and prepares the first brief; people validate sources, maintain assets, configure safe operating boundaries, approve consequential actions, and deliver the diagnosis or pilot to the customer.

Humans retain authority over consequential decisions. NERHIA does not sign contracts, commit prices, issue public alerts, close roads, activate infrastructure, or identify people. The AI prepares and explains the evidence; the operator validates the context and decides whether to act. This division makes the service useful while preserving accountability, especially when physical operations or public environments are involved.

The business model has three stages. First, an Urban Signal Diagnosis maps a bounded operational problem and recommends a pilot. Second, a 30-day pilot connects the relevant signals and delivers recurring command reports. Third, successful pilots can become NERHIA Pro subscriptions with an ongoing command console and operator support. Customer acquisition is intended to come from direct demonstrations, VibraAlto's commercial relationships, technology partnerships, and targeted outreach to operators of distributed equipment, facilities, vending fleets, and municipal innovation teams. Shared VBC infrastructure reduces duplicated engineering work across customers by reusing routing, evidence contracts, cost accounting, and safety controls.

During the eligible period, this distinct candidate business earned USD 0 in revenue, incurred USD 0 in attributed expenses, acquired 0 users, and had 0 paying users. We disclose those figures directly rather than treating interest, demonstrations, simulated checkouts, or technical readiness as traction. The project therefore does not claim product-market fit. Its present evidence is technical: a deployed public experience, repository history, automated tests, strict Gemini routing, canary controls, and an auditable execution path. The next commercial milestone is one independently paid, bounded diagnosis followed by a measured 30-day pilot.

The impact theory is that smaller operators will act faster and with greater accountability when every AI-generated recommendation carries its evidence boundary, uncertainty, and next verification step. Initial outputs will include normalized signals, generated briefs, provenance coverage, briefing latency, and human-reviewed recommendations. Commercial outcomes will include paid diagnoses, pilot conversions, recurring customers, retention, and the time operators save reconciling information. Safety outcomes will include the percentage of briefs that disclose uncertainty and the number of high-impact actions kept behind human approval.

Codex supported the build by inspecting the NERHIA and VBC repositories, translating the contest requirements into an evidence plan, implementing and reviewing strict Gemini routing, strengthening idempotency and cost traceability, hardening the redacted canary artifact, and organizing changes through reviewable pull requests. Codex did not accept rules, run paid infrastructure without authorization, invent financial evidence, contact customers, or send the final entry. Those decisions remain with León Canales.

Built With

• codex
• express.js
• firebase-hosting
• firestore
• gemini-api
• github-actions
• google-cloud-run
• node.js
• vbc-compute-layer