AI tool transforms access to UKWIR research

Access to critical water research has just got easier with the launch of a free, bespoke AI search tool from UK Water Industry Research (UKWIR).

The initiative supports the recommendations of the Independent Water Commission’s (IWC) final report, highlighting the crucial role of collaborative, evidence-based research to help guide the sector through this time of immense change.

UKWIR’s new AI search function provides organisations and stakeholders with rapid, cutting-edge research to inform policy and investment decisions.

Its two distinct functions are:

Contextual search: Users can ask questions in a conversational manner, for example – ‘What’s the best method of removing coliforms from drinking water?’ – and receive a list of relevant reports with specific page references.

Generative response: Acting like a ‘ChatGPT for water research’, the search tool pulls together information from multiple reports to provide concise summaries. Users can also specify the level of detail required, for example, asking for lists of pros and cons, or prioritised methods.

UKWIR’s team believes this could be a world-first for water research, and is sharing the findings, progress and learnings with the Global Water Research Coalition (GWRC), of which it is a member.

“At this pivotal moment, UKWIR is calling for even greater collaboration across the entire water sector, including its supply chain, and leading academic institutions,” said Mike Rose, UKWIR chief executive.

“It is extremely positive that the Independent Water Commission’s report directly identified UKWIR as playing a ‘key role’ in driving collaboration between the industry, research institutions, and academia. This endorsement underscores the vital role we play in supporting innovation and knowledge transfer.

“As the provider of impartial, science-based data, UKWIR stands ready to work closely with regulators, government, private businesses, and all other interested parties to ensure that policy and investment decisions are informed by the best available evidence, leading to a resilient, sustainable, and trusted water future for all.”

Rapid access to vital research

Historically, navigating the wealth of water research UKWIR has available online could be time-consuming, particularly when rapid, informed decision-making is required. The project uses artificial intelligence (AI) with large language models to speed up the process and make it easier to find information.

The project was spearheaded by UKWIR’s office manager Carol Ham, along with UKWIR’s research and communications co-ordinator Freya Caldwell.

“We put user experience at the heart of the design process when creating our new AI search function,” explained Caldwell. “A truly effective research platform needed to be intuitive, accessible, and capable of delivering robust and highly relevant results quickly, especially as the sector grapples with significant challenges and opportunities.”

“This free tool allows all stakeholders, members and non-members alike, to engage with UKWIR’s cutting-edge water research more efficiently. Just as importantly, users need to trust the information generated – which is why our source materials are always linked to provide complete transparency and traceability,” added Ham.

This transparency is crucial for ensuring that policy and investment decisions are built on a solid foundation of reliable data.

‘ChatGPT’ for the water sector

The bespoke AI search tool was created in collaboration with web developer Webree. It sits on the publication search page of UKWIR’s website, and unlike general-purpose AI models like ChatGPT or DeepSeek, UKWIR’s tool is hosted onsite and trained exclusively on UKWIR’s own library of reports and tools.

The tool uses a highly customisable Llama large language model, chosen for its prevalence in scientific research, to deliver tailored results for the water sector.

UKWIR also launched a redesigned website earlier this year, with an easy-to-use interface.

Previous articleUsing real-time monitoring and AI to clean up our rivers
Next articlePlugging the data gap for accuracy