Back office ops · Production

Totogi automates BSS change request processing from 7 days to hours using Amazon Bedrock multi-agent framework

The problem

Telecom BSS stacks consist of hundreds of applications from disparate vendors that are difficult and costly to integrate, relying on scarce specialized engineers. Each change request required thorough impact analysis and multiple rounds of coding, testing, and reconfiguration, while development cycles for new features often took months.

Workflow diagram · grounded in source
1
Change request initiated
trigger
“Multi-agent framework for fully automated change requests (CR), which reduces CR processing time from 7 days to a few hours”
2
Business analysis agent
ai_action
“The Business Analyst agent uses Claude's natural language understanding capabilities to process statement of work (SOW) documents and acceptance criteria. It extracts key requirements using custom prompt templates optimized for telecom B…”
3
Technical architect agent
ai_action
“The Technical Architect agent transforms business requirements into concrete AWS service configurations and architectural patterns. It generates comprehensive API specifications and data models and incorporates AWS Well-Architected princ…”
4
Developer agent code generation
ai_action
“The Developer agent converts technical specifications into implementation code using Claude's advanced code generation capabilities. It produces robust, production-ready code that includes proper error handling and logging mechanisms. Th…”
5
QA agent validation
validation
“QA agent validates the code for adherence to best practices, improving quality and security. It provides feedback which is used by the developer agent to update the code.”
6
Tester agent feedback loop
feedback_loop
“Tester agent generates robust unit test cases, streamlining validation and deployment. The result of the test cases is used by the developer agent to improve the code.”
7
Change request deployed
output
“This integrated pipeline reduces the time to complete a change request from 7 days to a few hours, with minimal human intervention”
Reported outcome

The BSS Magic multi-agent framework reduced change request processing from seven days to a few hours, and the automated testing framework achieved 76% code coverage, consistently delivering high-quality telecom-grade code with minimal human intervention.

Reported metrics
Change request processing timefrom seven days to a few hours
Automated test code coverage76%
Reported stack
Amazon BedrockAWS Step FunctionsAWS LambdaAnthropic ClaudeRAG
Source
https://aws.amazon.com/blogs/machine-learning/how-totogi-automated-change-request-processing-with-totogi-bss-magic-and-amazon-bedrock?tag=soumet-20
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

The BSS Magic multi-agent framework reduced change request processing from seven days to a few hours, and the automated testing framework achieved 76% code coverage, consistently delivering high-quality telecom-grade…

What tools did this team use?

Amazon Bedrock, AWS Step Functions, AWS Lambda, Anthropic Claude, RAG.

What results were reported?

Change request processing time: from seven days to a few hours; Automated test code coverage: 76% (source-reported, not independently verified).

How is this back office ops AI workflow structured?

Change request initiated → Business analysis agent → Technical architect agent → Developer agent code generation → QA agent validation → Tester agent feedback loop → Change request deployed.

Related back office ops cases