Achēv: AI Fraud Detection Unlocks Language Assessment Excellence

NGO Seeks Fraud Detection Services for Language Assessments

Our client, Achēv, is a not-for-profit organization that provides language learning services to for new Canadians. Their goal with these efforts is to support newcomers adapting to life in Canada. As part of these initiatives, the organization sought to develop an AI-based alternative for manual fraud detection services in their language assessments. Their goal was to ensure the integrity of language assessments, thereby preventing any malpractice or dishonest behavior during the assessment process.

Leveraging Amazon Rekognition for Secure Fraud Detection

To enhance the integrity of their language assessments, Achēv partnered with Adastra, supported by AWS, to address and overcome the challenges associated with language proficiency evaluation and remote service delivery to users with limited English and French skills. Adastra’s implementation of AWS fraud detection and test content delivery tools, guided by best practices, provided Achēv tailored solutions to meet their requirements effectively.

Through their collaboration with Adastra and AWS, Achēv discovered innovative language assessment fraud detection capabilities. The client received a robust solution, leveraging Amazon Rekognition capabilities, to identify common fraudulent activities during tests, such as unauthorized individuals in the testing environment, screen switching, and off-screen referencing.

This technology ensures the integrity of Achēv’s assessments, upholding standards and fostering trust in the evaluation process.

Robust Fraud Detection, Content Moderation and Cost Savings

Aided by AWS’s fraud detection features like face detection and screen monitoring, Achēv anticipates serving 20,000 clients annually. Their new fraud detection solution, powered by Amazon Rekognition, provides Achēv with the following benefits:

• Core backup and restoration services.
• Advanced analysis and recognition of unauthorized individuals.
• Content moderation to detect the use of unauthorized aids.
• Cost-effective, tailored solution with seamless integration.

Through the partnership with Adastra and AWS, Achēv gained the ability to conduct simultaneous assessments of multiple clients, significantly enhancing their operational efficiency and service reach. This advancement means that Achēv can now accommodate more language proficiency evaluations at once, streamlining the process and reducing wait times for users, thereby improving the overall accessibility and scalability of their services.

Moreover, the robust fraud detection capabilities provided by Amazon Rekognition transformed the security and reliability of Achēv’s language assessments. By leveraging this technology, Achēv can now automatically detect and address common fraudulent activities during tests, such as the presence of unauthorized individuals in the testing environment, screen switching, and off-screen referencing. This not only ensures the integrity of each assessment but also strengthens the trust and confidence of stakeholders in the evaluation process, establishing Achēv’s commitment to maintaining high standards and fairness.