What challenges did Revionics face before AI implementation?

Teams needed to build multi-agent systems for complex workflows without fragile, fragmented infrastructure. Revionics needed to automate pricing decisions while enforcing business logic such as competitiveness and margin constraints. Renault Group needed to prioritize EV infrastructure investments using multiple data sources with less manual effort. Nippon Television Holdings needed a maintainable production runtime for a video analysis agent and wanted to reduce development effort.

What was the impact of the AI solution?

Revionics reports the agentic setup automates pricing workflows while keeping business rules and forecasting considerations in the loop. Renault Group says the agent helps maximize driver convenience while reducing strain on its teams. Nippon Television Holdings says Agent Engine saved an estimated month of development time and simplified maintainability. Google positions the combined ADK and Agent Engine stack as a direct path from prototype to production for enterprise-grade agents.

Rank#337of 3,107

Source

GCP

Revionics and Renault Group multi-agent systems built with Vertex AI Agent Development Kit (ADK) / Agent Engine

Use case typeAutomotive operations automationUpdated Jun 13, 2026

Google Cloud introduced Vertex AI enhancements for building and operating multi-agent systems, including the open-source Agent Development Kit (ADK), the managed Vertex AI Agent Engine runtime, Google Agentspace integration, and support for Model Context Protocol (MCP). The article highlights real customer implementations: Revionics is using ADK to coordinate specialized agents and tools for pricing workflows, Renault Group is using ADK to help analysts prioritize EV charger installation sites using geographic, zoning, and traffic data, and Nippon Television Holdings is using Agent Engine as the backbone of a video analysis agent.

Organization: Revionics
Industry: Automotive
Location: Japan
Published: April 2025

Reported outcomes

Strategic outcomes

New product / capabilityAutomated pricing workflows with business rulesBetter decisions & insightPrioritized EV charger locations using multiple data sourcesSpeed & agilityReduced development effort for video analysis agentNew product / capabilityEnabled a maintainable production runtime

Primary read

Use case focus

Showing 3 of 4

1Agent orchestration
2Workflow automation
3Decision support

Teams needed to build multi-agent systems for complex workflows without fragile, fragmented infrastructure.
Revionics needed to automate pricing decisions while enforcing business logic such as competitiveness and margin constraints.
Renault Group needed to prioritize EV infrastructure investments using multiple data sources with less manual effort.
Nippon Television Holdings needed a maintainable production runtime for a video analysis agent and wanted to reduce development effort.

Revionics used ADK to build a multi-agent system that transfers between specialized agents and tools, combines pricing AI with agentic workflows, and reasons over large data artifacts rather than relying only on LLM context.
Renault Group used the ADK to develop an agent that helps data analysts leverage geographical, zoning, and traffic data to prioritize EV charger locations.
Nippon Television Holdings implemented Agent Engine as the production backbone for its Gemini-powered video analysis agent.
The platform also provides guardrails, orchestration controls, evaluation, monitoring, scaling, security, MCP connectivity, and integration with Google Agentspace.

Technologies

Vertex AI Vertex AI Agent Development Kit Vertex AI Agent Engine Gemini Google Agentspace Model Context Protocol

Revionics reports the agentic setup automates pricing workflows while keeping business rules and forecasting considerations in the loop.
Renault Group says the agent helps maximize driver convenience while reducing strain on its teams.
Nippon Television Holdings says Agent Engine saved an estimated month of development time and simplified maintainability.
Google positions the combined ADK and Agent Engine stack as a direct path from prototype to production for enterprise-grade agents.

Architecture

Vertex AI provides the production AI platform; ADK is used to design agents with deterministic guardrails, orchestration controls, and tool connectivity; Agent Engine is the managed runtime for deployment with context management, scaling, security, evaluation, and monitoring. The stack supports MCP-based connections to enterprise data and APIs, direct API and connector integrations, and optional registration with Google Agentspace.

Sources & evidence1

Groundedness: 4/5

Primary source

AI-generated summary. Verify important details with the linked sources before relying on this case.

Explore related AI use cases

Customer evidence