Accenture on 20 February announced the launch of new services for testing Artificial Intelligence (AI) systems globally which are powered with a “Teach and Evaluation” methodology intended to help companies build, track and quantify reliable AI systems within their own infrastructure or even in the Cloud.
Accenture Logo. Loic Venance/AFP
“As businesses adopt AI, it is vital to find better ways to train and maintain these systems — securely and with quality — to avoid adverse impacts on business performance, brand recognition, compliance and people,” Bhaskar Ghosh, Group Chief Executive, Accenture Technology Services, said in a statement.
Accenture’s “Educate and Evaluation” methodology ensures AI systems are producing the right decisions in two phases.
The “Teach” phase focuses on the choice of data, models and algorithms that are utilized to educate Machine Learning (ML).
Throughout the “Evaluation” stage, AI system outputs are compared to key performance indicators and evaluated for whether the machine can describe the way the decision or result was decided by using advanced techniques and Cloud-based tools to monitor the system.
“There is also a need for new capacities for evaluating data and learning models, picking algorithms, and monitoring for bias and ethical and regulatory compliance,” added Kishore Durg, Senior Managing Director, Growth and Strategy, and International Testing Services Lead to Accenture.
He said that Accenture’s “Teach and Test” methodology requires all of this into consideration to assist companies develop and confirm AI systems with assurance.
Accenture utilized this methodology to prepare a conversational virtual agent for a financial services firm’s website.