Regardless of sector, size of business or nature of services, revenue leakage (incorrect billing and collection of revenue and expenditure) is a common challenge. Due to a combination of contract complexity often involving thousands of documents (contracts and invoices), complex business models with extensive contractual and billing systems coupled with hundreds of different promotions, discounts, and modified pricing, poor administrative practices and a reliance on manual auditing practice, it is widely accepted that losses from leakage can impact an organisation by up to 10% of turnover reducing EBITDA by up to 25% and providing greater opportunity for fraudulent activity. In the case of the Telecomms sector for example, \> 1/3 of all operators estimate that at least 1% of total revenues is lost due to leakage -- costing the industry more than $23bn p.a. Despite being recognised as a critical business challenge; current best practice relies almost exclusively on manual practices by in house financial departments or through the engagement of external costly consultancies --based on contract sampling with an average of 0.01% of contracts audited. Attempts to automate the process through ERP systems, billing software, data aggregation and more recently AI based analytics still rely on manual input to extract the required data from customer contracts, Statement of Work (SOW's) and pricing amendments, with limited accuracy and scalability particularly across more complex service models/markets Cognitiv+ aim to address this market gap through the development of DeepRev -- a unique solution that combines advanced techniques in natural language processing (NLP) for clause mining (both text and figures), machine and deep learning with vision AI that will fully automate payment assurance management practices by accurately matching payments with contractual clauses to ensure compliance in real time. With application across all industry sector regardless of contractual and billing complexity, the solution will reduce leakage (by \>60% which equates to $2.8m for a company with a £100m revenue) with an immediate saving of 85% of manual costs and delivering \>90% accuracy. It also offers an improved means of reducing fraud and improve contract compliance with the potential to address wider challenges including Payroll leakage. With market need validated, a 18-month programme of Industrial Research is required to deliver a prototype which will be validated in a user trial. If successful, this combined functionality has the potential to truly disrupt the Revenue Assurance market with future opportunity to improve hospital management, PPP contracts.

"Companies are constantly challenged by the management of the ever-increasing contractual obligations, from payment schedules, delivery plans, to regulatory and compliance reporting. These tasks are currently done manually, going through every single contract, trying to detect the obligations by hand and somehow organising this information to be able to manage it. Currently, the only alternative to manual in-house contract management is professional service companies (lawyers or consultancies). Unfortunately, they also use the same methodology which means that they address the problem in a manual, non-standardised manner, allowing ample room for automation using innovative solutions. This project proposes the automation of the obligation extraction task using artificial intelligence, especially machine learning, natural language processing and smart contracts. Since the extracted information will be already in machine readable format we also propose the development of software the implements the workflows that have to do with obligation management(e.g. payment calendars, reporting notifications,etc.). The solution, accessed directly or as a service, will help legal, commercial and compliance professionals to accelerate contract review and analysis as well as avoid manual data entry into corporate systems allowing them to focus on higher value tasks. It will generate significant cost savings (50-90%) through the reduction of the time spent on manual tasks. On top of that, we expect significant operational risk reduction which will lead to reduced costs of litigation and potential penalties."

Legal Business Contracts govern the business relationship between trading business partners. They are like blueprints of expected business behaviour of all the contracting parties involved, and bind the parties to obligations that must be fulfilled by expected performance events. This highly innovative project proposes the automation of the obligation extraction task using artificial intelligence, especially machine learning and natural language processing. Since the extracted information will be already in machine readable format we also propose the development of software the implements the workflows that have to do with obligation management(e.g.payment calendars, reporting notifications,etc.). The solution, accessed directly or as a service, will help legal, commercial and compliance professionals to accelerate contract review and analysis as well as avoid manual data entry into corporate systems, allowing them to focus on higher-value tasks. It will generate significant cost savings (50-90%) through the reduction of the time spend on manual tasks. On top of that, we expect significant operational risk reduction which will lead to reduced costs of litigation and potential penalties.