Communicating via text chat is already a trend.

Customers prefer doing business with organizations via their favorite ways of communication.

ccR-Web automates repetitive customer inquiries via web chat.

For sizable organization, ccRobot improves overall custome care metrics. E.g. # of calls handled, call duration, calls waiting, # of free agents and etc.

Even for small business, 24x7 customer service via web chat becomes a feasible option similar to their larger competitors.

Product Name
Designed for
Customer Care Automation with Natural User Interface
Implementation Time
Approximately 2-4 weeks for most case without back office integration


  • Low Risk + Low Cost with our innovative migration strategy.

    ccRobot learns and improves as it is trained on the job by a live agent .

    Once ccRobot passes your bar exam, it will start working independently to serve repetitive inquiries.

    Unrepetitive inquiries will be escalated to live "supervisor". This experience can then be accumulated to grow ccRobot smarter.

    As a result, cost is lowered considerably for both initial deployment and on-going maintenance.

    In summary, customer care supervisor can train and improve the robot without expensive programmers.

  • High Return in terms of on-going cost savings.

    Fraction of budget to handle large amount of repetitive inquiries.

    Reduce the hiring, and training cost for high turnover rate.

  • High Quality

    Avoiding emotional agents

    Consistently high services quality

    Scalable customer care workforce for spikes in call volume

Major Implementation Milestones

  • Ground Work

    Gather talents E.g. IT Service Desk specialists

    Gather information E.g. FAQs, service agent training manual and historical call records.

    Build the Dialog Flow, if necessary

    Load some dialog examples (the seed training data)

  • On-the-job Training

    Service agent answers customer inquiries via operator console to record the conversation between callers and agents. I.e. callers’ requests and agent’s responses

    ccRobot will be trained nightly and will grow smarter.

    Refine Dialog Decision Tree Model as necessary.

  • Production Stage

    ccRobot will be released to production to answer repetitive inquiries

    ccRobot will continue to improve in production stage everytime the call is escalated to live agent.

Comparing ccR-Web with Traditional Chatbots

Traditional Chatbots ccR-Web
Technology Artificial Intelligence (AI) Natural Intelligence (NI)
Trained by Linguistics Specialists and Programmers General Business Users
Improvement Process Chatbot starts serving live customers after it is fully tested;
Subsequent improvement is made by programmers for the next release.
With increasing cost of each improvement, rate of improvement decreases over time; It is very difficult to be perfect.
ccR-Web starts serving live customers as soon as initial training sample such as FAQs is loaded;
ccR-Web is then trained on-the-job as an apprentice and grows smarter everyday.
At a much lower cost, continuous improvement over time with accuracy approaching perfect.
Self-Assessment N/A ccR-Web has built-in self-assessment capability that enables escalation process similar to those for human agents.
This serves as a safety net that never disappoints your customers.
Language Dependence Language dependent;
Each language has to be programmed separately by linguistics specialists.
Language independent;
ccR-Web support almost any natural languages and even mixed languages.