The Challenge

The authority’s brokerage team is responsible for assessing potential providers and determining which organisation is most suitable to meet an individual’s referral requirements. This process plays an important role in ensuring appropriate placements and maintaining service quality. 

 

The evaluation process relied on reviewing several documents produced by both internal teams and external providers. Staff needed to extract relevant information from referral documentation and compare this with responses received from providers. 

 

Several operational difficulties were identified. 

 

The process required considerable time and attention because staff needed to read large volumes of information across multiple documents. Important details were often distributed across several sources which made comparison difficult. 

 

Variation in interpretation could also affect consistency. Different reviewers could reach different conclusions depending on how they interpreted the documentation and assessed suitability. 

 

The process therefore placed a high administrative burden on the brokerage team and limited the number of cases that could be reviewed efficiently. As demand for services increased the council recognised the need for a more structured and scalable approach. 

Team collaborating using Databricks Genie to access governed data insights

The Approach

Telefónica Tech began the engagement with an AI envisioning workshop involving operational stakeholders and technology specialists. The workshop explored several potential use cases and assessed them using a scoring framework that considered impact, feasibility and alignment with organisational priorities. 

 

Provider evaluation emerged as a strong candidate because it involved repeatable processes and large volumes of documentation while also offering clear opportunities for improved efficiency. 

 

Following the workshop a two-day hackathon was organised through Microsoft Azure AI funding. The hackathon focused on creating a working proof of value that demonstrated how AI could support the evaluation workflow.

The Solution

The prototype solution introduced an AI assisted evaluation process designed to support staff rather than replace their expertise. 

 

Users can begin the evaluation by entering a reference identifier linked to the case. The system retrieves relevant documentation including referral forms and provider submissions. 

 

AI models then analyse the content of these documents and extract key information related to service requirements, capabilities and constraints. 

 

The solution compares provider responses against the requirements defined in the referral documentation. Based on this comparison the system produces a structured assessment that includes a score and a summary explanation. 

 

The output provides a ranked view of provider suitability which helps reviewers identify the most appropriate options more quickly. 

The interface presents this information in a clear summary which allows staff to review the AI generated insights and apply their professional judgement before making the final decision. 

The Outcome

The proof of value demonstrated how AI can support complex document analysis and comparison tasks, helping operational teams review provider submissions more efficiently and consistently.

Faster review of provider submissions

The solution reduced the time required for teams to review and assess provider documentation.

More consistent evaluation framework

Automated document comparison introduced a structured and consistent approach to evaluating responses.

Clearer identification of gaps and options

Staff were able to spot gaps in submissions more easily and compare provider options in a clearer, more structured way.

More capacity for professional judgement

By reducing repetitive document analysis, the brokerage team gained more time to focus on cases requiring deeper professional expertise.

Next Steps

Following the successful proof of value the council is exploring how the solution can be integrated into its operational systems. 

 

Work is also underway to improve the council’s data platform so that future AI initiatives can be developed more efficiently. 

 

The engagement has also provided a foundation for identifying further opportunities where data science and AI could support service delivery within the authority. 

Explore our Case Studies

Stay informed

Stay updated and subscribe to our regular communications