IBM’s Watson may replace pesky salesmen and telemarketers

by ruchi

No matter how good a particular product or service is, if its USPs aren’t conveyed in the best possible way, it can fail in the market. We have tele-marketers and onsite sales personnel to do that job and since there they are humans, their levels of proficiency are going to be different. IBM wanted to change this and hence is going to pass the baton of explain a product to a potential buyer to its much loved creation called Watson. Be it a huge retail outlet or support departments of companies, you could have Watson machines staged to take care of customer enquiries and concerns.

It is actually huge search engine that will answer questions and is powered by Big Blue’s DeepQA software. It has to be fed with loads of raw data i.e. general knowledge for Jeopardy, symptoms for medical diagnosis, or product specifications for sales and support and the software will turn it into useful, actionable facts- it perform analytical actions on the data and answer your queries. Also DeepQA has leverage over others for it is exceptionally good at understanding the natural language. The possibilities with Watson are endless. Even though right now is designed to augment and assist humans, but in future it will be able to function without human intervention too. The other application of this would be Twitter integration where users can tweet their questions and DeepQA will be more than happy to help.
The only hiccup for its wide-scale implementation is that both the hardware and software is incredibly expensive, both to purchase and to run. It has as much as 10 server racks filled with Power 750 servers, with a total of 2,880 processor cores and 15 terabytes of RAM. Though it’s difficult to comprehend its exact cost, but a little bit of industry info tells us that a 32-core Power 750 server costs around $350,000 and Watson uses 90 of these servers, let alone the installation cost of around $32 million. Woot!

Leave a comment