About the client
A UK-based financial consultancy specialising in investment and wealth advisory services. They advise some of the UK’s largest institutional clients and are supported by billions in assets under advice.
Business challenge
Manual and fragmented workflows slowing down and complicating annual reporting cycles
Our client operates in a highly regulated financial environment, where accuracy, consistency, and auditability are critical. Each year, their team produces hundreds of annual regulatory reports. These reports form an essential part of their client deliverables and are subject to internal reviews as well as external audit scrutiny.
However, the process relied heavily on manual workflows. Teams worked across Excel models and multiple source files, stitching information together through copy-paste and manual inputs. While functional, this approach made the process increasingly difficult to manage at scale. Key challenges included:
- Fragmented workflows: Data was spread across Excel sheets, documents, and source files with no single source of truth
- Manual effort and inefficiency: Repetitive data entry and copy-paste increased turnaround times
- Risk of human error: Multiple manual touchpoints created inconsistencies in calculations and reporting
- Complex review cycles: The team works under a multi-step review and approval process, and managing this workflow without a central system made tracking progress difficult
- Audit burden: Recurring auditor queries required significant time and consistent responses across clients
As reporting cycles repeated each year, these challenges compounded, making it harder for the team to maintain efficiency and control over the process.
Project goals
Our client was looking for a way to streamline and automate their end-to-end reporting lifecycle. With that in mind, GoodCore proposed an AI-powered reporting platform to process documents and automate time-intensive workflows. Our idea would:
- Capture and manage source documents and user inputs in a structured, centralised system
- Use AI-driven data extraction to convert unstructured inputs into structured data, and generate report sections based on defined numbers, thresholds, and business rules
- Enforce a role-based, multi-step review and approval workflow with full audit trail and visibility
- Leverage AI to generate contextual responses to auditor queries, supported by a repository of standardised answers and source data
The solution
The system is designed to integrate with the client’s existing reporting methodology, while enhancing it with intelligent automation. By combining workflow management with AI-driven data processing and content generation, the platform reduces manual effort and automates the end-to-end reporting lifecycle.
AI-driven report generation from complex inputs
The platform uses GPT-4o to transform complex inputs from varied source documents into standardised, regulator-ready reports.
Input and source document management
Users can upload and manage all source inputs, including accounting data and supporting documents, which are linked directly to specific report sections.
Excel model integration
The system integrates with the client’s Excel-based model as the core calculation engine, allowing outputs to be extracted and mapped accurately.
AI-powered document processing
The AI model interprets different Excel formats and document layouts, extracts relevant data and generates structured outputs aligned with predefined report templates and business logic.
AI-generated commentary
The system generates contextual narratives, including explanations of key movements, comparisons with prior periods, and supporting notes.
The Solution
AI-powered auditor query response generation
The platform uses natural language processing (NLP) to automate the handling of auditor queries. By analysing source documents and historical responses, the system generates contextual, consistent, and audit-ready answers.
AI-generated responses: The system automatically drafts responses to auditor queries based on relevant source data and prior inputs
Context-aware generation: AI tailors each response using underlying data and document context, rather than relying on generic answers
Template-backed consistency: A repository of standard responses to common queries ensures uniformity across clients and reporting cycles
Human-in-the-loop control: The client’s team can review, edit, and approve responses before sharing, ensuring accuracy and accountability
The impact
- The platform is now used by 20+ team members across preparation, review, and approval roles, creating a single, shared workspace for reporting.
- 400+ reports generated annually through the system, replacing manual, Excel-driven report creation.
- Report preparation time reduced by 40–50%, allowing teams to handle more reporting cycles without increasing headcount.
- Consistent workflows and validations have ensured high accuracy and compliance standards across all reports.
- Automation enabled the business to increase reporting capacity by 2–3x, taking on more engagements and directly contributing to revenue growth.
Looking to create an impact with software?
Explore how we can work together.