Technology Provider: SmartAction

Client Company: J&B Medical Supply

Client Company Description: J&B is a medical supplier based in Michigan, offering a diverse selection of products from over 950 manufacturers (including diabetics, urological, ostomy, wound care, and more)

Industries: Healthcare

Applications: Text and Language Processing


J&B Medical’s expansion goals were being hindered by a heavy need to recruit and train live call center agents. The hiring process for this role was particularly time-consuming and costly. Automation efforts for low-to-medium complexity calls was limited by the elderly demographic of J&B Medical and HIPAA compliance. HIPAA’s authentication rules (requiring three pieces of customer data) was particularly time consuming and mundane, but much of the process hadn’t been able to be automated.

Actions Taken:

SmartActions worked with J&B Medical to model and mimic the best practices of live agents through natural language processing and other artificial intelligence processes. SmartAction complied with HIPAA regulations by developing a system for garnering three pieces of customer information to positively identify the caller. The system was designed to provide multiple options of authentication if the customer (or their caretaker) did not have the information at hand.


A few short weeks after going live, customers were engaging with the self-service on 96% of calls, and 70% of those callers were fully authenticated without being transferred to a live agent. SmartAction claims that self-service authentication took less than half the time it took agents, reducing handle times by over two full minutes. By pulling approximately 50,000 agent minutes out of the call center, J&B experienced cost savings of around 65% over their live agents.

Transferable Lessons:

When people hear “NLP” (natural language processing) today, they often think about devices like Apple’s Siri, or of a cutting-edge chatbot system. Call center automation may not seem glamorous, but it’s safe to say that, at present, call center automation drives much more efficiency and dollars to the bottom line than chatbots. Its likely to stay that way for many years, as companies tinker with how to make chatbots work well with consumers.

The lessons from this case study, however, can be extended outside of call centers. Nearly any repeated, mundane, relatively predictable conversational task (whether it be via email, phone, or chat) is low-hanging fruit for artificial intelligence. For any organization that must offer round-the-clock support and ensure compliance and consistency, voice services of this kind might be a useful model for AI application (this is especially pertinent in HIPPA and/or PCI compliance requirements). SmartAction claims that these systems also support the training of new agents by modeling best-practices precisely.


TechEmergence conducts direct interviews and consensus analysis with leading experts in machine learning and artificial intelligence. Stay ahead with of the industry with charts, figures, and insights from our unparalleled network, including executives from Facebook, Google, Baidu, Yahoo!, MIT, Stanford and beyond: