6.8X FASTER DISCOVERY FOR A LEADING INVESTMENT COMPANY

A coordinated multi-agent discovery workflow accelerated the cycle from 34.2 to 5 days, reduced development cost by 84%, and reached a 6.8x acceleration factor.

INTRODUCTION

Discovery becomes difficult when business requirements are fragmented across multiple sources and critical knowledge lives in undocumented legacy systems.

A leading investment company faced this challenge in a complex modernization initiative. Requirements were distributed across tools including Confluence, Jira, Figma, and legacy code, while the target solution also depended on an undocumented legacy system and several external teams.

CI&T introduced a coordinated discovery workflow with seven specialized AI agents covering reverse engineering, requirements consolidation, dependency mapping, and story preparation.

Powered by CI&T FLOW, the model brought multiple information sources into a connected process, improved requirement quality, and kept people responsible for orchestration and decision-making.

The initiative achieved a 6.8x acceleration factor, reduced the discovery cycle from 34.2 to 5 days, and lowered development cost from $18K to $2,915.

THE CHALLENGE: CONNECTING FRAGMENTED REQUIREMENTS

Requirements were distributed across multiple tools and systems, making it difficult to create a complete and reliable view before development.

Fragmented Information

The team needed to deliver 46 stories across 10 epics within a compressed delivery schedule.

Legacy Knowledge Gaps

Critical functionality depended on an undocumented legacy system that needed to be understood before it could be integrated into a modernized platform.

External Dependencies

Multiple external dependencies and non-agentic ways of working added complexity to the discovery and decision-making process.

THE SOLUTION

CI&T deployed seven specialized AI agents to consolidate information, reverse-engineer legacy systems, map dependencies, and prepare requirements for development.

Requirements Discovery

Agents combined business and technical information from multiple sources to create a more complete and consistent view of requirements.

Legacy Analysis

Reverse-engineering capabilities helped uncover undocumented system behavior and connect legacy knowledge to modernization needs.

Story Preparation

The workflow optimized story splitting, creation, and dependency mapping, helping teams move from discovery to actionable development inputs faster.

THE IMPACT

The coordinated discovery model accelerated decision-making, improved requirement quality, and reduced the effort required to prepare work for development.

FASTER DEVELOPMENT

Acceleration factor
reached 6.8x

Cycle reduced
from 34.2 to 5 days

85% faster
discovery cycle

LOWER COST

Development cost
reduced by 84%

From $18K
to $2,915

Cost governance enabled
at epic level

STRONGER DECION-MAKING

Faster business
refinement

Improved requirement
quality

Humans remained responsible for orchestration

TECHNOLOGIES USED

CI&T
FLOW

CI&T FLOW Enterprise AI Management System supported the coordinated use of specialized agents across discovery, analysis, and story preparation.

MULTI-AGENT
DISCOVERY

A coordinated workflow of seven specialized agents supported reverse engineering, requirements consolidation, dependency mapping, and story creation.

LEGACY
ANALYSIS

AI-assisted reverse engineering helped uncover undocumented system behavior and improve the context available for modernization decisions.