recentpopularlog in

AI

« earlier   
Ways to think about machine learning — Benedict Evans
"Returning to the parallels with relational databases, this might be rather like talking about SQL in 1980 - how do you get from explaining table joins to thinking about Salesforce.com? It's all very well to say 'this lets you ask these new kinds of questions', but it isn't always very obvious what questions."
ai  ml  trends 
5 hours ago by davewsmith
How computers could make your customer-service calls more human • WSJ
Daniela Hernandez and Jennifer Strong:
<p>Cogito is one of several companies developing analytics tools that give agents feedback about how conversations with customers are going. Its software measures in real time the tone of an agent’s voice, their speech rate, and how much each person is talking, according to Dr. Place. “We measure the conversational dance,” he says.

That dance is sometimes out of sync, such as when an agent speaks too quickly or too much, cuts a customer off, has extended periods of silence or sounds tired.

When the software detects these mistakes, a notification pops up on a window on an agent’s screen to coax them to change their strategy. The alerts are useful not just for the agents, but also for their supervisors, Cogito says.

When insurer MetLife Inc. started testing the software about nine months ago, Emily Baker, a 39-year-old supervisor at a call center in Warwick, R.I., thought: “Why do I need this artificial intelligence to allow me to be more human? How much more human can I be?”

But the program has come in handy when coaching new agents, she says, especially those with little experience. One of her 14 agents said the software noticed he wasn’t speaking with enough energy, so it prompted him with a message to pep up plus a coffee-cup icon, she says.

Tiredness can come off as lack of confidence, Ms. Baker says, and it’s important for clients to “feel confident about the service we’re providing” because callers are often going through potentially life-changing events. The call center where Ms. Baker works is focused on disability insurance.</p>


Machines to watch over us, and correct us when we aren't good enough with each other.
callcentre  cogito  ai  software 
yesterday by charlesarthur
Metro - Data Crowdsourcing for AI and Machine Learning >>> Generate and sell your own data
About:
"Anybody can use our community-made plugins, "DataSources", to generate data and sell it in real-time to a Project. (free for now) A Project could be a groundbreaking new AI, a research study, or someone's curiosity project.
[...]
Artificial Intelligence is fueled by data, so innovation is monopolized by those who control data.
[...]
Our philosophy is "Empower, Don't Restrict". Let users control and monetize their data.
[...]
We want to create a transparent and fair system of sharing any kind of data where both the producer of data and the buyer of data have the power they deserve. Right now, the middle-man has all of the power and that is why the world is dominated by Google and Facebook. This is a fair and honest system which will be built without imposing unreasonable restrictions upon what can and cannot be done.
[...]
We are not a data collection company. We're a data energizing company.
[...]
We have created a Chrome/FF extension to which users can attach "DataSources" - simple community-made pieces of code which allow users to generate their own data. Our website hosts "Projects", which are pieces of work which want to crowdsource data from users using specific DataSources. DataSources are made and curated by the community, not by us, and they are open source so anybody can verify them. This gives the community direct access to raw data with its intrinsic innovative value for the coming AI revolution, while giving individual users direct access to the monetary value."

HAG
Startupinthecloud  Agrachina  AI  machine-learning  Data  Platform  open-data  opensource  Crowdfunding  Business_Advisory_relevant  Business_Model  Business  Project_Information_Platform  community  Marketing  UrB 
yesterday by eocas

Copy this bookmark:





to read