Hope, award winner of the Summit for Action on AI’s call for projects

HOPE, or Harmonizing Optimization and Precision in Emissions, is a carbon impact calculation application used by consultants in their projects and employed by sales teams.

It goes beyond merely focusing on the reduction of carbon emissions; it also offers a simulation model that allows the quantification and understanding of the environmental impact of the company's actions. This is a highly effective communication and awareness tool that, through practice, strengthens Talan’s commitment to responsible technologies.

The scope of the project includes the integration of AI into the already operational HOPE tool, initially serving as an eco-calculator, and the expansion of its capabilities to automate processes, make informed predictions, and provide actionable recommendations. This integration enables Talan to reinforce its sustainability efforts while responding to client demands for transparent and comprehensive carbon footprint assessments.

The "Hope" project is a striking example of how Talan is addressing the environmental challenges we face. By using an application to simulate greenhouse gas emissions during consultants' assignments, this project highlights a technology that is both responsible and human-centric.

This solution falls within the scope of the augmented consultant theme.

The HOPE project is designed to align with Talan's sustainability goals and the expectations of its clients. By integrating AI into the eco-calculator, the project not only aims to strengthen Talan’s CSR commitments but also serves as a model for the use of AI to promote sustainability in the consulting and technology sectors.

Hope is one of the 111 winners of the "Efficiency for AI" call for projects launched by the Summit for Action on Artificial Intelligence (February 10-11, 2025).

Can AI and CSR coexist harmoniously?

The project primarily impacts Talan’s CSR initiatives and its operations with clients, particularly those in the public sector who require detailed carbon footprint assessments. The enhanced HOPE tool also supports business development functions by offering a competitive edge in responding to tenders with sustainability requirements. Furthermore, the recommendation system can influence operational planning by suggesting optimised resource use and travel arrangements.

Why is AI crucial for this project?

AI is key to achieving these objectives as it significantly reduces the manual effort and expertise required to calculate carbon impacts and explore alternatives. The use of AI improves both the efficiency and accuracy of the tool, while uncovering carbon reduction opportunities that might otherwise be overlooked.

The data used for this project includes a dataset of tender offers and commercial proposals. Additionally, manual annotations are available for these documents, specifying which information can be directly extracted and what can be inferred from the text. This annotated dataset provides a critical foundation for training and evaluating the AI models used in the project.

How is AI being utilised?

The implementation of AI focuses on two main approaches. The first uses large language models (LLMs) for traditional information extraction tasks, while the second involves a mini-Retrieval-Augmented Generation (mini-RAG) methodology. The mini-RAG approach combines a lightweight retrieval mechanism with generation capabilities to achieve accurate results efficiently in terms of resources. These models are designed to complement the existing capabilities of the eco-calculator by automating data extraction and providing intelligent insights.