YDUQS & CI&T

How YDUQS modernized its educational hub operations through development with Artificial Intelligence

OVERVIEW

YDUQS, a renowned technology and services group in higher education in Brazil, faces the ongoing challenge of modernizing and optimizing its operations at semi-presential educational hubs.

To meet this demand, the partnership between CI&T and YDUQS's Information Technology team developed the "Polos" project, aimed at modernizing processes and enhancing user engagement and satisfaction.

THE CHALLENGE

01.

Manual
Processes

YDUQS's operations relied on manual execution processes that, in light of business growth, presented opportunities for automation. The goal was to convert these stages into digital workflows to ensure greater agility, scalability, and accuracy in both hub management and the student experience.

02.

Decentralization of Information

The lack of a centralized system created friction in accessing information, compromising the agility of operations and regulatory compliance.

03.

Regulatory Compliance

YDUQS needed solutions that not only optimized processes but also ensured that all operations were in compliance with industry regulations.

THE SOLUTION

Al-First Culture and Immediate Efficiency

The first wave of the project focused on implementing an Al-First culture. We began with a comprehensive mapping of the Software Development Life Cycle (SDLC) and cultural appropriation by the development team.

• Refinement: Implementation of Planning agents for the automatic creation of User Stories.
• Quality: Assurance of High Clean Architecture, ensuring quality standards.
• Development: Generation of contextualized code aligned with architectural planning.

To measure the impact, we established a standardized metric using Business Complexity Points (BCPs), creating a historical baseline of 7.1 BCPs/FTE.

Spec-Driven Development (SDD) with AWS's Kiro

The second wave represented a leap in productivity as we implemented Spec-Driven Development (SDD) with AWS's Kiro tool. The choice of the tool was made in partnership with YDUQS's Information Technology team, which played a crucial role in evaluating and directing the solution most aligned with business needs and the company's technological strategy.

This approach reverses the logic of traditional development: instead of the team writing thousands of lines of code, they focus on creating rich and detailed technical specifications that serve as an accurate "blueprint" for the solution.

Kiro's AI interprets these specifications and automatically generates the application code. As a result, the team shifted its focus to design and strategy rather than manual coding. The outcome was transformative: a complex functionality, originally estimated to take a month, was developed, tested, and delivered in just one week.

THE RESULTS

Impact on Productivity

• First Wave: The integration of AI into the SDLC elevated the average quarterly productivity to 37.8 BCPs/FTE, a 5.2-fold increase compared to the baseline.

• Second Wave: Projected productivity jumped to 50 BCPs/FTE, achieving an efficiency 7.1 times greater than the baseline.

• Culture and Scalability: The new delivery methodology established a replicable operational standard for other areas of the company, transforming efficiency into a cultural asset and ensuring sustainable corporate gains in the long term.

Note: BCP (Business Complexity Points) is an open-source functional scope measurement scale from CI&T that helps measure the throughput of software produced.

Technical Results

• Speed, Reusability, and Quality: AI accelerated Backend development while maintaining high standards of technical excellence and ensuring complete consistency in code pattern replication.

• Living Documentation: The automatic generation of technical documentation increased the autonomy of the team.

• Empowerment of the Team: Developers were able to focus on business rules, taking on a more strategic role.

TECHNOLOGIES USED

Node.js: For Agile and Efficient Backend Development
React: For creating interactive user interfaces.
CI&T Flow: For optimizing development processes.
Kiro AWS: For implementing Spec-Driven Development.
AWS: Robust infrastructure to support the developed solutions.