Titulo Estágio
AI-Driven Software Development Optimization
Áreas de especialidade
Engenharia de Software
Sistemas Inteligentes
Local do Estágio
Hibrido (1 dia por semana em Vila Nova de Gaia) e 4 dias em formato remoto
Enquadramento
This research proposal presents a comprehensive plan to investigate the intersection of artificial intelligence (AI) and software development, with a primary focus on the study and application of Generative AI (GenAI) – specifically leveraging ChatGPT and Microsoft Copilot. The overarching goal is to revolutionize the software development lifecycle by enhancing efficiency, agility, and innovation through GenAI technologies.
The proposed work plan unfolds across four strategic work packages, each building upon the findings and advancements of the previous phases.
Work Package 1: AI Integration in Software Development Processes
This phase concentrates on analyzing trends in open-source repositories using TensorFlow, scikit-learn, and natural language processing (NLP) tools. The goal is to identify areas where GenAI can significantly impact software development workflows. Additionally, the study evaluates GenAI's influence on agile frameworks and its application in continuous development methodologies.
Work Package 2: GenAI Automation Methodologies in Software Development
The second work package dives into the development of GenAI-enhanced scripts for user stories, automation of technical documentation, and integration of GenAI in automated testing and sample code generation. By seamlessly integrating GenAI tools like ChatGPT and Copilot, this phase aims to streamline and optimize key software development processes.
Work Package 3: Integration of GenAI in Demonstrator Development
Focusing on practical applications, this package explores GenAI's role in prototype development across diverse fields. From energy sector optimization to life sciences, telecommunications, and automotive software development, the objective is to showcase tangible benefits and real-world implementations of GenAI across different industries.
Work Package 4: Practical Application of GenAI in Diverse Fields
The climax of the research involves the application of GenAI in energy, life sciences, telecommunications, and automotive sectors. Predictive maintenance, renewable energy integration, accelerated drug discovery, autonomous vehicle code generation, and predictive maintenance for complex systems are among the varied use cases demonstrating the transformative potential of GenAI in addressing industry-specific challenges.
Objetivo
The proposed research aligns with the objective of optimizing software development through the strategic integration of GenAI. The findings from each work package contribute to a holistic understanding of GenAI's impact on diverse aspects of the software development lifecycle. This research not only advances theoretical knowledge but also provides practical solutions with the potential to revolutionize industries, making GenAI an integral component of future software development practices.
Plano de Trabalhos - Semestre 1
Work package 1 with the following tasks:
1. AI-Driven Trend Analysis for Software Development.
▪ Utilize TensorFlow and scikit-learn to analyze trends in open-source repositories, focusing on areas where GenAI can enhance software development.
▪ Extract insights from articles and technical documentation using natural language processing (NLP) with NLTK, specifically addressing improvements in software de-velopment workflows.
▪ Develop scalable solutions to handle large datasets and diverse sources, emphasiz-ing GenAI integration.
▪ Provide actionable insights for optimizing software development through the ap-plication of GenAI.
▪ Complete the trend analysis within the specified timeframe, emphasizing the im-pact of GenAI.
2. Evaluating GenAI Impact on Agile Development Frameworks.
▪ Benchmark Scrum and Kanban frameworks to identify GenAI's role in optimizing agile practices in open-source projects.
▪ Evaluate the effectiveness of GenAI-enhanced agile practices in projects using stat-ic code analysis tools like SonarQube.
▪ Conduct a comprehensive analysis of agile methodologies, emphasizing the inte-gration of GenAI for improved code quality.
▪ Enhance project management methodologies through the integration of GenAI in agile software development.
▪ Complete the benchmarking and analysis within the allocated time, showcasing GenAI's impact on agile frameworks.
3. Applying GenAI in Continuous Development Methodologies.
▪ Investigate the application of GenAI in continuous integration and continuous de-livery (CI/CD) practices using Jenkins and GitLab CI.
▪ Analyze the automation of tests with GenAI-driven tools like Selenium and JUnit to enhance code quality and development speed.
▪ Implement GenAI-powered CI/CD pipelines for open-source projects, ensuring seamless integration and delivery.
▪ Improve the efficiency and reliability of software development processes through the application of GenAI.
▪ Complete the assessment and implementation of GenAI-driven CI/CD practices within the established timeline.
Work package 2 with the following tasks:
1. Developing GenAI-Enhanced Scripts for User Stories.
▪ Implement Python scripts using GenAI, specifically ChatGPT and Microsoft Copilot, to automate the generation of user stories based on project requirements.
▪ Utilize GenAI frameworks like ChatGPT and Copilot to enhance script development for user stories in an open-source environment.
▪ Develop scripts that seamlessly integrate with GenAI, aligning with project re-quirements and optimizing user stories creation.
▪ Improve efficiency in user stories creation through the application of GenAI-driven automation.
▪ Complete the GenAI-enhanced script development within the assigned timeframe for user stories.
2. GenAI-Driven Automation of Technical, Functional, and Code Documentation.
▪ Implement a documentation pipeline using GenAI tools like ChatGPT and Copilot for automatic generation of technical and functional documentation from the source code.
▪ Integrate GenAI tools to extract documentation directly from code, ensuring con-sistency between documentation and codebase.
▪ Create a streamlined process for generating comprehensive documentation through the application of GenAI.
▪ Enhance documentation processes by leveraging the capabilities of ChatGPT and Copilot.
▪ Implement and optimize the documentation pipeline within the specified time, showcasing GenAI's impact on documentation.
3. GenAI in Automated Testing and Sample Code Generation.
▪ Utilize GenAI frameworks like ChatGPT and JUnit for creating automated tests and generating sample code.
▪ Ensure the consistency between documentation and implementation by automati-cally generating code samples using GenAI.
▪ Develop a comprehensive suite of automated tests and sample code that reflects the capabilities of GenAI in software development.
▪ Ensure the reliability and correctness of software through the application of GenAI in the testing and code generation processes.
▪ Complete the implementation of GenAI-driven automated tests and code genera-tion within the allocated time.
Plano de Trabalhos - Semestre 2
Work package 3 with the following tasks:
1. GenAI-Enhanced Demonstrators in Python and Flask.
▪ Develop prototypes of demonstrators using GenAI frameworks, focusing on ChatGPT and Microsoft Copilot, to enhance software development capabilities in Python and Flask.
▪ Utilize GenAI capabilities for efficient data manipulation in the demonstrators, showcasing the synergy between AI and software development technologies.
▪ Create functional prototypes that highlight the practical applications of GenAI in API development using Python and Flask.
▪ Demonstrate the tangible benefits of integrating GenAI in real-world software de-velopment scenarios.
▪ Complete the development of GenAI-enhanced Python and Flask demonstrators within the specified timeframe.
2. GenAI Integration with NoSQL Databases in Demonstrators.
▪ Utilize GenAI, particularly ChatGPT and Copilot, in conjunction with NoSQL data-bases like MongoDB and Cassandra to enhance data storage capabilities in demon-strators.
▪ Implement efficient queries using GenAI-driven tools for analysis and performance optimization in NoSQL databases.
▪ Showcase the seamless integration of GenAI with NoSQL databases, emphasizing flexibility and scalability.
▪ Demonstrate the compatibility and benefits of GenAI in real-world projects with diverse data storage requirements.
▪ Complete the integration and optimization of NoSQL databases with GenAI within the established timeline.
3. GenAI-Powered User Interface Development with React and Material-UI.
▪ Create modern and responsive user interfaces using React and Material-UI, en-hanced by GenAI capabilities such as ChatGPT and Copilot.
▪ Integrate GenAI-powered interfaces with demonstrators, ensuring a consistent and intuitive user experience.
▪ Develop interfaces that align with best practices in UI/UX design, leveraging the capabilities of GenAI.
▪ Showcase the seamless integration of GenAI, specifically ChatGPT and Copilot, in demonstrator projects to enhance user interface development.
▪ Complete the development of GenAI-powered React and Material-UI interfaces within the assigned time.
Work package 4 with the following tasks:
1. Energy Sector Optimization.
▪ Objective: Apply GenAI to optimize software development processes in the ener-gy sector, focusing on efficiency improvements, predictive maintenance, and re-newable energy integration.
▪ Use Case 1: Predictive Maintenance for Power Grids.
▪ Description: Implement GenAI-driven algorithms to analyze historical data and predict potential failures in power grid components.
▪ Practical Application: Enhance reliability and reduce downtime by proactively addressing maintenance needs, ultimately improving energy grid perfor-mance.
▪ Use Case 2: Renewable Energy Integration
▪ Description: Utilize GenAI to develop algorithms that facilitate the seamless integration of renewable energy sources into existing energy grids.
▪ Practical Application: Optimize energy production, reduce dependency on non-renewable sources, and contribute to sustainable energy practices..
2. Advancements in Life Sciences Software Development.
▪ Objective: Explore GenAI applications in life sciences to accelerate software devel-opment processes related to drug discovery, genomic analysis, and healthcare so-lutions.
▪ Use Case 3: Accelerated Drug Discovery
▪ Description: Apply GenAI to automate the generation of code for analyzing large datasets in drug discovery, speeding up the identification of potential drug candidates.
▪ Practical Application: Expedite the drug discovery pipeline, leading to faster development of pharmaceutical solutions.
3. Telecommunications Infrastructure Enhancement.
▪ Objective: Explore GenAI applications in life sciences to accelerate software devel-opment processes related to drug discovery, genomic analysis, and healthcare so-lutions.
▪ Use Case 3: Accelerated Drug Discovery
▪ Description: Apply GenAI to automate the generation of code for analyzing large datasets in drug discovery, speeding up the identification of potential drug candidates.
▪ Practical Application: Expedite the drug discovery pipeline, leading to faster development of pharmaceutical solutions.
4. Automotive Software Development Revolution
▪ Objective: Revolutionize automotive software development with GenAI, focusing on autonomous vehicle systems, predictive maintenance, and software-defined vehicle features.
▪ Use Case 5: Autonomous Vehicle Code Generation
▪ Description: Utilize GenAI to automatically generate code for autonomous ve-hicle systems, including perception, decision-making, and control.
▪ Practical Application: Accelerate the development of autonomous vehicles, ensuring robust and efficient software implementations.
▪ Use Case 6: Predictive Maintenance for Automotive Systems
▪ Description: Apply GenAI to predict potential failures in automotive systems by analyzing sensor data and automating the generation of maintenance rou-tines.
▪ Practical Application: Improve vehicle reliability, red
Condições
The student will be paid 1.018,52 euros per month for a period of up to 12 months.
Observações
Supervised by Dr. Marco Araujo and prof. Paulo Macedo.
Orientador
Marco Araújo
marco.araujo@capgemini.com 📩