Case Study

Reducing Audit Risks in Tax Credit Submissions

Crater Labs developed a neural model to highlight potential risk factors in R&D tax credit submissions, saving hundreds of hours in audit defence times.

The Client

A boutique consulting firm providing R&D tax credit preparation and advisory services to leading Canadian technology companies. This model allows the firm to automate quality assurance process, and quickly identify audit risk factors.

The Challenge

R&D tax credit preparation is a complex field, consisting of project narratives, costing reports and tax information.

In a fast-paced environment, our client needed software that could streamline quality assurance processes by assessing thousands of variables in a tax filing to highlight potential audit risks.

The Solution

Crater Labs developed a model combining text/entity analysis in combination with financial data in tax returns to identify potential audit risks associated with R&D tax credit filings.

This project benefited our client in several ways. It reduced the time to conduct quality assurance, better communicate addressing of risks to the client and also reduce audit preparation times, by identifying what documentation to prepare pro-actively. With this investment, our client was able to save hundreds of hours in claim and audit preparation time within a year.

IP Generated

We implemented an ensemble model consisting of an autoencoder to perform natural language processing and a CNN to examine the relationships between the thousands of fields on a tax return. The model uses this complex data to predict factors that can lead to an audit.

Benefits & ROI

Our solution allowed our client to reduce the time to conduct quality assurance, and pro-actively identify corrective measures with its clients. By identifying audit-risk factors ahead of time, our client was able save hundreds of hours in audit preparation time and provide an unparalleled level of service in a highly competitive field.

Keep Reading

More Moonshots Worth Celebrating

Minimizing radiation exposure in construction projects

We developed a custom LSTM neural network to predict radiation exposure of construction workers in sensitive sites, improving shift scheduling and safety compliance.

Identifying Potential Security Access Violations

We developed a semi-supervised neural network to identify anomalous security access card swipes in a large corporation with numerous facilities throughout the world, previously hidden secure access violations.

Reducing Audit Risks in Tax Filings with Deep Learning

Crater Labs developed a neural network to highlight potential risk factors in R&D tax credit submissions, saving hundreds of hours in audit defence times.

See All Case Studies