What challenges did Rentokil face before AI implementation?

Technicians had to dig through siloed systems and old documents before site visits to find customer and site history such as past pest identifications, door codes, and recommendations. The manual process created administrative burden and slowed service readiness.

What was the impact of the AI solution?

Developed a functional AI proof-of-concept within six months. Pilot participants reported a significant reduction in the time needed to gather relevant information. Technicians favored the hands-free audio interface and feedback identified and resolved issues within hours or a day. Rentokil plans to scale the application to 25,000 field technicians in North America and later other regions.

Source

GCPProductionEvidence: Medium55/100

Rentokil & Telana deploy a Gemini-powered hands-free assistant for field technicians

Use case typeField service automationUpdated 1 day ago

Rentokil partnered with Telana to build On The Go, a Gemini-powered assistant for field technicians that gives hands-free, real-time access to customer and site history through internal APIs. The solution uses Gemini Live for bidirectional audio streaming, Firebase for frontend services, Gemini Enterprise Agent Platform with an ADK agent, and regional Cloud Run for low-latency scaling. The pilot in Buffalo was developed in about six months and focused on reducing manual research before site visits, improving technician preparedness, and capturing rapid field feedback.

Organization: Rentokil
Industry: Professional Services
Location: United Kingdom
Published: July 2026

Reported outcomes

Strategic outcomes

Employee experienceHands-free technician experienceCost efficiencyReduced manual research before site visitsSpeed & agilityRapid pilot development and issue resolution

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Rentokil
Provider: GCP
Maturity: Production
Linked source: Google Cloud Customer Stories

Used Firebase Hosting, Authentication, and Functions; deployed the ADK agent to Gemini Enterprise Agent Platform; and ran streaming communications through regional Cloud Run

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 2 of 2

1Field service automation
2Conversational assistants

Technicians had to dig through siloed systems and old documents before site visits to find customer and site history such as past pest identifications, door codes, and recommendations.
The manual process created administrative burden and slowed service readiness.

Built OTG as a generative AI assistant with conversational access to internal customer data using internal APIs.
Implemented Gemini Live bidirectional streaming for hands-free interactions while technicians are driving.
Used Firebase Hosting, Authentication, and Functions; deployed the ADK agent to Gemini Enterprise Agent Platform; and ran streaming communications through regional Cloud Run.

Technologies

Gemini Gemini Live API Gemini Enterprise Agent Platform Firebase Cloud Run Agent Development Kit

Developed a functional AI proof-of-concept within six months.
Pilot participants reported a significant reduction in the time needed to gather relevant information.
Technicians favored the hands-free audio interface and feedback identified and resolved issues within hours or a day.
Rentokil plans to scale the application to 25,000 field technicians in North America and later other regions.

Architecture

A hands-free field technician assistant built with Gemini Live bidirectional audio streaming, internal APIs, Firebase Hosting/Authentication/Functions, Gemini Enterprise Agent Platform with an ADK agent, and regional Cloud Run for low-latency streaming and scaling.

Implementation partners1

Telana

Sources & evidence1

Evidence: Medium55/100Evidence strength