Deep-dive

Take a closer look at what went into delivering this solution.

Discovery

The initial discovery phase revolved around understanding the complexities of contract management in the construction industry, particularly with JCT contracts. Stakeholders expressed frustration with the ambiguity surrounding contract clauses, and the unclear assignment of responsibilities within the contracts, especially when real-world changes on-site meant a perceived change in designation. As the team delved into the issue, it became apparent that the construction industry lacked an efficient, tailored solution to automate and streamline contract review and clarify roles and risks. This discovery set the stage for ideation, transforming the problem into an opportunity for an innovative AI-powered assistant that could expedite contract analysis and reduce potential legal and operational risks.

Challenges

Throughout the development process, several key challenges arose. One of the primary hurdles was ensuring that the solution could accurately interpret complex construction contracts, often filled with jargon and dense legal language. Additionally, we faced challenges related to data privacy and security, as we were working with sensitive contractual information. Ensuring the solution adhered to strict industry standards for privacy was non-negotiable, thus a privatised in-house solution rather than utilising cloud-based systems was adopted. Another challenge was the integration of the Retrieval-Augmented Generation (RAG) technology, which required careful fine-tuning to ensure that the assistant could not only pull relevant data from a wide range of contract templates but also provide accurate and actionable insights. Lastly, aligning the system with the specific needs of different stakeholders, contract managers, legal teams, and project managers required continuous iteration and feedback loops.

Use of AI

The solution leverages advanced AI, particularly natural language processing (NLP) and Retrieval-Augmented Generation (RAG) technology, to solve the problems identified during the discovery phase. NLP was applied to analyze the intricate details of JCT contracts, identify potential gaps in coverage, and detect ambiguities in responsibilities. RAG technology allowed the system to pull in relevant examples and similar case studies, helping to highlight potential risks and inconsistencies within the contract. By processing vast amounts of contract data in real-time, the AI assistant could offer precise suggestions on clause amendments, role clarifications, and risk mitigation strategies, all tailored to the specific needs of the client, within the bounds of the construction industry. The secure deployment of this AI solution ensured that all sensitive data was protected, while the user-friendly interface allowed stakeholders to easily interact with the assistant and receive actionable insights, all while reducing manual contract review time and increasing overall efficiency.