Skip to content Skip to sidebar Skip to footer

Source: Silicon.fr as of 01-04-2020

 

In an en increasingly world, the branches  and business departments are asking CIOs to “abandon”  the support function posture that meets a need, to  become  a  true  “business partner” offering  innovative solutions, to  get ahead of  their  needs..

It is  important to note  that  this  change  in   posture  must  take into account  the budgetary constraints  of the ISD,  i.e.,at constant  scope to  reduce the budgetary  share  allocated  to  the maintenance (MCO), ITDs  must  “do   better  with as much.” 

In this context, it is  essential for CIOs  to find  and implement all possible levers  possibles  to ensure an   ever better and constant budget service  offering.  

 en  Within recurrent   operational maintenance  (MCO) charges, incident management is  a  major  part  because  it   consumes resources    and  is generally  poorly  optimized..

What is  an  incident and “incident  management”?

According to the  definition  of the Information Technology Infrastructure Library  (ITIL),  an incident  refers to  “Any  event that is not  part  of the standard operation of a service and that causes,  or  may  cause,  an  interruption  or decrease in the  quality  of  these  services. and in  this  context,incident management  is  characterized  by a  goal that aims to “restore as quickly as  possible the normal functioning  of services,  the  negative impact on business activities  and  thus ensure that the  highest  levels  of  quality  of service and  availability  are  maintained”  

 Reading this   definition,it en is clear that the process and  tooling  have  changed little  since  the introduction of IT services. .

Today,following the  finding of an incident by a  user,(customer, a trade correspondent,   other  person…) The incident  is  declared  by hand in the  form of a ticket   allowing   it  to enter  the  resolution process. .

Then,  this  ticket  is  sent  to a  technician in charge  of carrying out  an  initial  analysis of the situation  in order to   understand the cause to  redirect  this  ticket  to the technical  resolution  team  (dev,  network,…) who  seems    most  competent  to  solve the anomaly..

This very  expensive process en  in  “man  time”   is no  longer  optimized  adapted in a world  where  man,is getting more and more help, by machines and applications  ever  more  efficient  and  fast. It  is becoming  clear  that  a  transition  to  AI-based robotics    to  optimize  certain low-value-added     tasks  needs  to  take place..

How to optimize  this  process  with artificial intelligence

The concept of artificial intelligence  was born in the 1950s,  and   consists of implementing    a  number  of techniques to enable  machines  to  imitate  a  form of real intelligence. .

In strong  development  since  the 2000s,  there  is  no  doubt  that  Artificial Intelligence  is  now  an important lever in  optimising the costs  of MCO and the  quality  of services of the IS,  including   a  automation    of the ticket creation part  and  incident analysis. analyse

This trend  is   constantly  increasing thanks  to the constant evolution  of  AI as it  moves  towards  a so-called  “strong”  AI that will be able to learn  on its own  based on its  own  experience.

The power of the intervention  of a  “strong”  AI analyzing the unstructured  fields  of the base of incidents / tickets and  interconnecting with the server logs   or  network,  allows  to make  a  correlation  of the symptoms reassembled  and  thus   define the accidental points of the      applications  studied  or  the IS  analysed  to   optimize  the redirection of tickets  to  the right  interlocutors (i.e.  those  with  capacity   of  solve  the incident)without a flat reduce  the  number  of callers  and  thus  the time to process  incidents  or in the longer  term to preventive  maintenance  in advance.

In addition, thanks to the en  exponential development of deep learning  (an in-depth  and  automatic  learning method  based  on  learning data models), the  efficiency  of  the analysis  done by the machine will be  constantly  increasing  so  that after a certain period of time,  no human  intervention  will be  necessary  in the  resolution  of incidents.
 This  method   automates    up to 80% of a company’s    tickets/incidents,    which  would allow men to be repositioned  on higher value-added tasks.

Add   to  this that machines work faster, 24/7 and with a  much lower rate  of  failures  and  errors  than   humans, the productivity gains  that  can  be  induced  by  the use  of  AI are     not  negligible  and  significantly increase  the  quality of IS services with an  improvement  in  the  quality of IS services of up  to  40% (according to  a  study  conducted  by  the  US  economic  and financial news channel  CNBC).

This deployment  of  AI  is    already  engaged  in  the optimization  of CIOs, start-up   such as  HypeSense  have  embarked  on  this  promising market  segment. .

The AI today  is  in the acceleration phase,  each  year  the  cost  of   setting  up decreases,  while  the  ability to   understand,  learn,  compute.   It  therefore becomes  impossible not to be interested in the  subject and not to see it  as  a leading performance tool.  

With 67%(according to  a  study  by  TNP Consultants) of CIOs who  now think  that  AI  can help them  grow,the question  is  not  whether  we  should turn  to  this  new  assistant  but  when  Will we have to take the  plunge?

Authors – Amaury de Gonneville  (Senior  Consultant) and Guillaume  Savry  (Consultant)

Show CommentsClose Comments

Leave a comment

News ORS © 2020. All Rights Reserved.