How to create a great encyclopedia for the employee
Banking services in Russia are actively used by more than 50% of the population. Regardless of age, specialty and budget size, it is important for the client to receive complete and accurate information about the status of accounts and quick answers to their questions. For this, banks optimize telephony, develop conversation scripts, create robots to respond to typical requests. In addition to technology, advanced teaching methods are being introduced, regular testing of employees' knowledge and assessment of service quality is taking place, but it’s impossible to keep everything in mind. Especially if there are dozens of different products in the bank’s product line, there are promotions and special offers for customers.
Under the cut today we talk about our Wikipedia - a smart directory where there is all the information about all the products and services of the bank.
')
In the bank branches today you are met by young and energetic employees, whose average age is 25 years. They show empathy, strive for professional growth and value the manufacturability of processes. The key competence is the ability to quickly find the answer to any client question. Of course, if the employee is unable to fully advise the client, he will connect the senior shift or redirect the call to another service. But which client would like to expect another connection?
In the case of a telephone consultation, the wait time is a key SLA. Therefore, we undertook to create a single knowledge base with smart navigation.
Such reference books, of course, have been used before. But the teams developed them at different times for their project and according to their own understanding. As a result - technical inflexibility, complexity of integration and support.
For example, for one of the previously created solutions, a functionality was implemented that allowed saving only videos and presentations in pdf format. The information was placed on a temporary server and was available for a limited time, there was little server space, of course, there was little, there were no opportunities for targeted content placement, taking into account the structure of the department, positions and geography: everyone saw everything.
To unify the development of individual components, component teams were allocated.
Their task is to create universal modules that have common characteristics, an understandable use case, and can be easily integrated into a business initiative - teams of any direction: retail, corporate or CIB. One of these universal and necessary modules is the employee reference book. It should be noted that the use of such components by business teams significantly saves the resource for development and, most importantly, the correction of possible errors.
The approach to working in the component team is no different from business development: design thinking is a must for all initiatives. Participants identified key consumers, divided into groups, conducted interviews. Following the results of data analysis, Invision created a prototype and tested it on users.
The basis of the solution 2 thesis:
During the session of generating ideas, the participants proposed features that were subsequently cut into sprints for the team:
Development is carried out iteratively, now uses a version with limited functionality.
At the start of development, there was no accessible wiki engine that could provide support for a large volume of users and would be easily embedded in the existing technical circuit of the bank.
Taking into account the timing of the assembly, a solution from IBM Filenet was chosen as a temporary solution for the backup, which allows you to store documents and use an additional search engine. In fact, the current version of the knowledge base is a file repository, where documents are displayed as articles.
During the course of work, the team analyzed existing wiki engines: Confluene, IBM connections, Mediawiki, Xwiki. For the content and knowledge storage system, the following was viewed: wikipedia, imdb, web knowledge base on Apple, Samsung, Tesla and other sites. Free wiki engines such as mediawiki allow you to implement a solution as quickly as possible, but have limitations on integration and scalability.
Currently, the development is carried out in two directions: one team is engaged in integration with existing services, the second is working on the Presentation Layer (PL) of all applications, is leading the development of IT and service architecture and provides technical support for related development.
We decided to use Atlassian Confluence as a targeted backup solution. This will increase the reliability and scalability of the solution by an order of magnitude, taking into account the future number of users - more than 300,000, as well as significantly increase functionality due to additional plug-ins that can be purchased or created by yourself. Pleased with the content editor - one of the most advanced in terms of functionality, while it has a clear and simple interface. The wide range of services and APIs that the system provides simplifies integration with it.
Work continues, in future publications we will talk about the implementation of a targeted solution and the practice of fixing bugs.
Have you encountered the task of creating such a service?
The most popular programming languages used by engine developers are PHP, Java, Perl, C #, and Python. In our Java stack is the basis, so we suggest chatting in the comments and choosing the best wiki engine written in Java.
Under the cut today we talk about our Wikipedia - a smart directory where there is all the information about all the products and services of the bank.
')
In the bank branches today you are met by young and energetic employees, whose average age is 25 years. They show empathy, strive for professional growth and value the manufacturability of processes. The key competence is the ability to quickly find the answer to any client question. Of course, if the employee is unable to fully advise the client, he will connect the senior shift or redirect the call to another service. But which client would like to expect another connection?
In the case of a telephone consultation, the wait time is a key SLA. Therefore, we undertook to create a single knowledge base with smart navigation.
All at once
Such reference books, of course, have been used before. But the teams developed them at different times for their project and according to their own understanding. As a result - technical inflexibility, complexity of integration and support.
For example, for one of the previously created solutions, a functionality was implemented that allowed saving only videos and presentations in pdf format. The information was placed on a temporary server and was available for a limited time, there was little server space, of course, there was little, there were no opportunities for targeted content placement, taking into account the structure of the department, positions and geography: everyone saw everything.
Component teams
To unify the development of individual components, component teams were allocated.
Their task is to create universal modules that have common characteristics, an understandable use case, and can be easily integrated into a business initiative - teams of any direction: retail, corporate or CIB. One of these universal and necessary modules is the employee reference book. It should be noted that the use of such components by business teams significantly saves the resource for development and, most importantly, the correction of possible errors.
The approach to working in the component team is no different from business development: design thinking is a must for all initiatives. Participants identified key consumers, divided into groups, conducted interviews. Following the results of data analysis, Invision created a prototype and tested it on users.
The basis of the solution 2 thesis:
- data should be presented in the form of articles, not files in a directory, with advanced cognitive search;
- adaptability of functionality to the mobile platform.
During the session of generating ideas, the participants proposed features that were subsequently cut into sprints for the team:
- voice search;
- automatic solution search based on the analysis of the voice request of the client;
- internal rating and rating system;
- introduction of an emotional discharge simulator;
- adding a cross-reference "with this often look."
Development is carried out iteratively, now uses a version with limited functionality.
Temporary solution
At the start of development, there was no accessible wiki engine that could provide support for a large volume of users and would be easily embedded in the existing technical circuit of the bank.
Taking into account the timing of the assembly, a solution from IBM Filenet was chosen as a temporary solution for the backup, which allows you to store documents and use an additional search engine. In fact, the current version of the knowledge base is a file repository, where documents are displayed as articles.
During the course of work, the team analyzed existing wiki engines: Confluene, IBM connections, Mediawiki, Xwiki. For the content and knowledge storage system, the following was viewed: wikipedia, imdb, web knowledge base on Apple, Samsung, Tesla and other sites. Free wiki engines such as mediawiki allow you to implement a solution as quickly as possible, but have limitations on integration and scalability.
Next steps
Currently, the development is carried out in two directions: one team is engaged in integration with existing services, the second is working on the Presentation Layer (PL) of all applications, is leading the development of IT and service architecture and provides technical support for related development.
We decided to use Atlassian Confluence as a targeted backup solution. This will increase the reliability and scalability of the solution by an order of magnitude, taking into account the future number of users - more than 300,000, as well as significantly increase functionality due to additional plug-ins that can be purchased or created by yourself. Pleased with the content editor - one of the most advanced in terms of functionality, while it has a clear and simple interface. The wide range of services and APIs that the system provides simplifies integration with it.
Work continues, in future publications we will talk about the implementation of a targeted solution and the practice of fixing bugs.
Have you encountered the task of creating such a service?
The most popular programming languages used by engine developers are PHP, Java, Perl, C #, and Python. In our Java stack is the basis, so we suggest chatting in the comments and choosing the best wiki engine written in Java.
Source: https://habr.com/ru/post/352112/
All Articles