My five ideas: 

1.    A way for customers to view anticipated accurate wait time before ‘calling’ in; for example – when going to uplussorg.ca, I see the wait time in minutes and the suggested time to call in. I also see the capability to either A) talk to a chatbot or B) request a call back – both options which can further deflect by using AI to complete the case or at least get it started and triage accordingly. Bonus marks for triaging to provide dynamic wait-time and take into consideration extra capacity (e.g. if there’s a hospital 40 minutes away by helicopter that may be a better option for a brain surgery if all operating rooms are busy at the local hospital for the next 8 hours) – so take a hospital for example – if I list my symptoms, it will instruct me when: 1) when I should start driving to the hospital 2) which hospital I should drive to 3) things I can do in the meantime (e.g. CPR, connect a defibrillator, ice, elevation, rest, breathing into a bag, etc.)… or if the problem is serious enough it would send an ambulance based on severity/ priority of the current queues 

2.    A way for customers to request to be called rather than having to call in – for example Sue can hit a Form on the channel of her choice which tries to understand the problem first then create a case that goes into a queue where an agent will call her. The Amazon customer service model followed this paradigm with the added feature that a machine would first connect the next customer – i.e. call them and play an automated greeting: you are about to be transferred to discuss your problems A,B,C. Jessica at Amazon will help you, please press 1 to proceed to hear music until she’s ready which should be momentarily… using AI to perform the basic triage, we can ensure that the person calling back is the right resource (this is helpful in a financial institution where people are often bounced around between 2-3 human resources before getting to someone that can process the initial request). 

3.    Virtual planning box. It’s not uncommon for organizations like hospitals to operate where disparate stakeholders have conflicting goals and may infringe upon each other’s space. Rather than letting it spin out of control it can be better understood by using a group-based approach (e.g. Delphi approach) and having people vote on issues important to them. By looking at the consensus, this can help prioritize key issues for discussion and resolution either virtually or in short periodic meetings. It also separates the “group-opinion from the noisy-person’s-opinion”  Stakeholders who raised the issue and voted for it are engaged and informed as the organization addresses the concern. The groups have different agendas (for example in a hospital – Brain Surgeons vs. ENT Surgeons vs. Oncology Doctors vs. Breast Surgeons vs. Nurses vs. Radiological Doctors vs. Radiological Techs… etc. etc… these are all diverse groups with different needs and regulatory requirements… conflict amongst these groups can not only affect operational efficiency but also severely impact morale as people suffer through passionate arguments. 

4.    Surgery-bot – to be used during a surgical procedure. Can start with automating simple journeys – “help me call Dr. X with priority STAT” – could translate to a voice/video of the stakeholder in real time. Side note here….. – in “Do No Harm” by Henry Marsh he tells of a time when he waited over an hour to contact a fellow surgeon while, to say the least, it should have been done immediately given the severity. Marsh, in the midst of a surgery for the past 6 hours with a team of five other human resources. The skull was cut and open. Marsh ran into a problem, advanced imaging was required before he could proceed; Marsh needed to understand whether a particular mass was benign, which would impact the next steps of the surgery. To interpret the image, another Doctor (call him Doctor Y) was required to read and interpret the medical image. Doctor Y was on his lunch break and ignored notifications. After more than an hour of frustration, Dr. Marsh took off his surgical scrubs (“de-scrubbing”) and proceeded to locate a telephone outside the operating-room and call Doctor Y. He was frantic (hungry, annoyed, tired, angry) when he learned that Doctor Y was ignoring the high-priority request. After the emotional outburst, Marsh proceeded back to the operating room (which took another 20 minuets to get “scrubbed-in” where the patient lye with their brain open and exposed…………. Other ideas may include: “Delay surgery, 6 hours” – “myocardial infarction (MI) STAT” –… the idea is to provide a new way to do these things – not to override current procedures – i.e. the surgeon always vocally calls out for someone else to do the work – in this case we could have audio to listen, process with NLP, and help guide the assistant through his journey as he assists the surgeon… 

5.    A way to process employee requests and triage accordingly in a retail environment – can also be used for real time matters. Take the example where a TJX employee asked a patron to leave the store with their service-animal. Surly if we had a tool readily available for the employee, he could have used it to look up operational procedure… in reality he went with his gut and an onslaught ensued on social media, where patrons were outraged at the behavior of the brand. How can we avoid this? I personally worked in customer service on the floor in a retail context (Staples Business Depot) and we had communication devices and were instructed to first inquire if anything was beyond our scope. Perhaps we can mimic that same voice channel and provide a quick litmus test to the query – red/yellow/ green. The employee would articulate an action, and he would receive feedback in 300ms