TechSee releases ‘first’ computer vision platform for contact centers

Tel Aviv-based TechSee has unveiled what it says is the first AI-powered visual recognition for contact centers.

Called TechSee Smart, it builds on the company’s success with its previous TechSee Live. With that earlier service, contact center agents can see still images or video on a customer’s smartphone and make drawings for the customer on top of the visuals in an augmented reality layer. The firm says more than 100 client companies are currently using TechSee Live.

Analyzing the problem

The new service allows a smartphone-calling customer to send still images or video to the contact center, which are then processed by the TechSee Smart platform to determine the product model and the cause of the problem.

In a typical use case, the contact center’s interactive voice response (IVR) system sends a SMS link to the customer’s smartphone, while the customer is speaking with a live agent or while waiting for one.

Through the link, the customer sends several photos or video clips of the product having a problem. The TechSee Smart platform, which is integrated with Salesforce’s customer relationship management (CRM) system and several contact center platforms, visually analyzes the imagery, recognizes the product and determines the likely problem — assuming there are visual indications.

Chatbots can’t tell what’s wrong

For example, the platform could let the agent know that the product is model #ABC cable set-top box, and that one of the connecting wires is attached to the wrong connection. The TechSee Smart platform’s deep learning technology is trained with images and data relating to actual use cases in contact centers, primarily in the industries of telecommunications, consumer electronics and utilities.

VP of Product Marketing Liad Churchill told ClickZ that the processing of objects and the likely cause of the problem should take only a few seconds, once the images or video clips have been uploaded. He added that his company’s testing indicates TechSee Smart’s issue resolution is about 95% accurate.

The primary use case for TechSee Smart is when a user calls on the smartphone, and Churchill said about 90% of users calling into contact centers do own such a device. If the user communicates through a landline phone or a computer, the agent can also send a SMS link to the user’s smartphone.

Churchill added that, today, spending in customer service for AI is largely for text or data analysis, but “chatbots can’t tell you why your [product] isn’t working.”