What challenges did Fortinet face before AI implementation?

Complex and growing product documentation made it time-consuming for customers to find the right answers. The chatbot had to scale across many products and versions while meeting strict accuracy and trust requirements. Fortinet also needed to control inference cost and latency for large-scale customer support use.

What was the impact of the AI solution?

Fortinet reported an 85x reduction in inference costs compared with its previous model. The chatbot improved latency and customer experience. The solution reduced specialist and developer workload by filtering simple questions and routing deeper issues to experts. Fortinet is planning widespread public use of the chatbot across its customer base.

Source

Fortinet reduces cost and latency for customer support chatbot using Amazon Nova on Amazon Bedrock

Use case typeIT operationsUpdated Jun 13, 2026

Fortinet, a cybersecurity company serving more than 800,000 customers worldwide, wanted to make its complex product documentation easier to access through a generative AI support chatbot. The company needed a scalable solution that could handle more than 60 products and multiple documentation versions while maintaining high accuracy, trust, and low response time.

Organization: Fortinet
Industry: Tech & Comms
Location: United States
Published: May 2026

Reported outcomes

85x

costCost savings

Strategic outcomes

Cost efficiencyLowered inference costsSpeed & agilityImproved chatbot latencyCustomer experience & trustImproved customer support experienceScale & capacityScaled support across products and versions

Catalog median for cost savings deployments: −45% across 345 reported metrics. Compare benchmarks →

Primary read

Use case focus

Showing 3 of 3

1Customer Support Automation
2Knowledge Management
3Generative AI Chatbot

Complex and growing product documentation made it time-consuming for customers to find the right answers.
The chatbot had to scale across many products and versions while meeting strict accuracy and trust requirements.
Fortinet also needed to control inference cost and latency for large-scale customer support use.

Fortinet worked with AWS and the Generative AI Innovation Center to build a generative AI support chatbot using Amazon Bedrock.
The solution uses Amazon Bedrock Knowledge Bases to ground responses in Fortinet's private documentation sources.
Fortinet applied Amazon Bedrock Guardrails to align responses with responsible AI and application safeguards.
After evaluating multiple models, the company switched to Amazon Nova Micro on Amazon Bedrock to improve price performance and lower latency.

Technologies

Amazon Bedrock Amazon Nova Micro Amazon Bedrock Knowledge Bases Amazon Bedrock Guardrails

Fortinet reported an 85x reduction in inference costs compared with its previous model.
The chatbot improved latency and customer experience.
The solution reduced specialist and developer workload by filtering simple questions and routing deeper issues to experts.
Fortinet is planning widespread public use of the chatbot across its customer base.

Architecture

A generative AI support chatbot built on Amazon Bedrock, grounded with Amazon Bedrock Knowledge Bases connected to Fortinet documentation and protected with Amazon Bedrock Guardrails. The team evaluated foundation models and deployed Amazon Nova Micro as the chatbot model to optimize cost and latency.

Implementation partners1

Generative AI Innovation Center

Sources & evidence1

Groundedness: 5/5

Primary source

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

Explore related AI use cases

Customer evidence