Credit & Collections – Utilising Machine Learning

Credit & Collections – Utilising Machine Learning

Machine learning (ML) is a branch of artificial intelligence (AI) and computer science that focuses on using data and algorithms to imitate how humans learn, gradually improving its accuracy.

We have built our own collections learning system based on our analysis of over 130 million real-world customer and agent conversations since 2018. ML algorithms are based on stored data used to make decisions or predictions based on inbound customer conversations.

The initial implementation of ‘supervised learning’ is based on two aspects

1. Access to the Saascoms AI engine

2. An agent manually tags each conversation

When tagged, the contents are applied to the Saascoms ML function. The agent can then type a reply or select from a branded template, adding personal customer information into the response via a link to the organisation’s CRM.

When a response is registered within the Saascoms engine, it is added to the data pool for analysing and adding to the ML algorithm.

As the AI becomes more confident, we move to ‘semi-supervised’ learning, where the engine will identify a response to the agent. The agent then approves the content and sends it. The aim is to ensure that content is approved to maintain accuracy and compliance.

The final process of our ML is ‘unsupervised learning’.

Data is analysed and added to the algorithm without human intervention, thus becoming fully autonomous. This process can take several months and relies on stage 1 and 2 sign-off processes between Saascoms and the collection’s organisation.

Every conversation is allocated a confidence score. Depending on intent and sentiment, i.e. vulnerability, it must equal a 95% confidence match. If this criterion succeeds, the ML algorithm will trigger the workflow without human intervention. If the confidence score is under 50%, the ML algorithm is disengaged and passed to an agent.

Benefits to our clients.

• Compliance and accuracy
• Reduces human error, ensuring continuity of information and responses for all customers.
• Offering templated and consistent communications to customers.
• Full transcript reporting for chatbot, agent, and customer interactions, regardless of the communication channel, i.e. email, SMS, live chat, social
• Linking previous conversation threads to the new inbound activity, using the content to predict sentiment and intent, and routing the customer to the relevant team based on requirements.

Efficiencies

• Quicker responses reduce conversation wait times. On average, service level response times were cut by 93.6%.
• Out of hours 24/7/365 service. Saascoms Chatbots resolved 90,000 collection and arrears conversations in Q1 2022, outside the 8 a.m.– 9 p.m. hours.
• Available across multiple channels (SMS, email, Live Chat, and social, i.e., Facebook Messenger, Whatsapp) within the same Saascoms conversational platform – Omnireach.

Management Information

• Offers MI to Collection & Arrears clients based on when customers are engaging, which customers prefer to engage by channel, and when to contact customers to ensure the best engagement.
• Identifying vulnerable customers to ensure the correct contact strategy is used moving forward.

In conclusion, machine learning has become an essential tool in the credit and collections industry. By using data and algorithms to imitate how humans learn, machine learning is gradually improving its accuracy.

Saascoms has developed its own collections learning system based on the analysis of over 130 million real-world customer and agent conversations since 2018.

The system uses supervised, semi-supervised, and unsupervised learning to ensure compliance and accuracy, reduce human error, and offer templated and consistent communications to customers.

The system also provides quicker responses, out-of-hours 24/7/365 service, and management information that identifies vulnerable customers to ensure the correct contact strategy is used moving forward.

Saascoms’ technology has proven to benefit businesses by improving efficiencies, reducing wait times, and providing valuable insights into customer engagement.

Saascoms have a wealth of experience and are award-winning at that, so book a demo and see how our technology can benefit your business.

Contact Us