16 ways we use Watson now
Blog: Capgemini CTO Blog
According to a recent CB Insights analysis of earning calls transcripts, AI is the single “mega-trend” on corporate executives’ agenda. AI is opening up countless possibilities in many areas—augmenting employees, automating processes, transforming customer experience and decision-making, or shaping new businesses. You might be wondering where to start, or where to go next. As food for thought, here are some ways we have been using Watson for our customers in different industries.
- Product information chatbot. For an insurer, we have built a virtual agent complementing their customer services team to answer customer questions on insurance products, coverages, policies, etc.
- Customer services chatbot. For an online banking client, we integrated their chatbot with Salesforce to enable the chatbot to retrieve customer information and address most customer requests, with the objective of the chatbot becoming an effective cross sell engine
- Claims management agent assist. For insurers, we built solutions bringing insurance policy details and other enterprise documentation to the fingertips of call center agents handling claims, to help them reply to their customer’s detailed questions in real time.
- Startup watch. For a bank, we created a tool supporting automation of the examination of startup companies. The sentiment analysis capabilities of Watson were a key differentiator to accurately define whether a company was being viewed in positive terms.
- Audit calls automation. For a telco customer, we automated audit calls, making it possible to listen to 100% of the calls (vs. ~1%), measuring several compliance and quality measures, detecting fraud earlier and gathering new customer satisfaction insight.
- HR chatbot. We helped a customer build a chatbot answering the most typical HR questions.
- Shopping assistant. We built a smart conversational app that helps consumers with their online purchases.
- Retail automation. To address the key pain point in the shoe-buying process (finding a sales rep in the store, waiting for them to go check the inventory, etc.), we built an app recognizing shoe models, checking availability in the inventory and prompting a sales person to bring the right model to the customer.
Healthcare and Life Science
- Market research. For a life sciences player, we performed market analysis, identifying competitive information, market trends, and more.
- Treatment and medical research. For hospitals, we conducted multiple studies to improve treatment decisions based on analysis of years’ worth of anonymized patient files. Cases included: improving decision factors to perform imaging on children in ER; quality insurance on compliance with cancer treatment pathways; identification of cancer “companion diseases,” underestimated medication side effects and comorbidity factors.
Other services industries
- Voice of the customer. For a public service, we built a service crawling inputs in social media apps to gather valuable insight on improvement areas.
- Pepper robot customer service. For an airline, we are piloting the use of a robot to assist travelers at airports.
- Staffing. We are augmenting the capabilities of our staffing resources to identify the best resources for a project with a tool that reads resumes, projects descriptions, and identifies the best candidates throughout the organization.
- Service desk automation. Our service desk offering includes automation of typical call intents that can be handled by a chatbot.
Manufacturing and transportation
- Robotized product testing. Combining robotics, IoT, visual recognition and NLP technologies, we are supporting customers with robotized testing of cockpits and other control stations.
- Visual inspection. Our solution substitutes human inspection of typical quality defects (i.e. car paint) on production lines.
And much more.