Feedback tool with evaluation via AI

Feedback tool with evaluation via AI

RationalizationUnlock sales potentialPlug and Play
"Evaluation without AI would have taken a team of four about a week. By using AI, this was done in a matter of seconds."
Project costs (digits)4
Minimum purchase0
License byAccounts
SubscriptionYes
SSOn/a
API (% of features)n/a
Browser UIYes
Docker possibleNo

The problem

To get valuable feedback about the change from classroom to online teaching by COVID-19, a university lacked the possibility to obtain a questionnaire with free text to validate and quickly evaluate it

The university was forced by Corona at very short notice to switch from classroom teaching to online teaching. Afterwards, information was missing on what went well during the changeover and what problems might have occurred. Without a quick and flexible evaluation, the university would not have been able to assess on a quantitative basis whether the changeover had worked well and what measures should be taken next.

The solution

With a survey/evaluation tool, more than 10,000 mostly free-text answers could be quickly evaluated using AI - in conformity with the DSGVO

The couchsurvey.com team worked together with the university to evaluate the technical requirements for the platform and was able to implement the individual adjustments within one week. In the same time the questionnaire was designed in cooperation with the university. After the survey had been live for 2 weeks and more than 10,000 answers had been received, the results, which mainly consisted of free text answers, were analysed and visualised with the help of artificial intelligence, more precisely with procedures from the field of NLP. AI had the additional advantage of eliminating the subjectivity factor in the text analysis for the university.  

No other survey platform would have been able to implement the individual adjustments during this time. Additional benefit: The solution is one of the few DSGVO-compliant systems which does not have to be hosted by the university itself.

Illustrations

Before

  • Uncertainty about how well the transition to digital teaching has worked
  • Uncertainty when creating and designing surveys
  • Little time capacity for data evaluation
  • No clearly defined procedure 

Afterwards

  • Low time and cost input due to the cooperation
  • Clear requirements for a survey
  • Clearly defined and fast procedure
  • A satisfactory survey with meaningful results that did not have to be evaluated manually, but were automatically analyzed, evaluated and clustered by the system

Insights

Stumbling blocks

The university was suddenly confronted with the challenge of evaluating many thousands of free text answers. An artificial intelligence for the evaluation  was developed so that the university did not have to evaluate the data manually. It was possible to do this mainly because of the proximity to the customer and the fast communication.

What had the customer tried before?

Question Pro, Limesurvey, Unipark, Evasys

Additional challenges at the customer

Fast action and 100% privacy compliance.

Differentiation from other solutions - in general

  • DSGVO compliant cloud solution (Plug&Play incl. DSGVO)
  • State-of-the-art technology and design
  • Possibility of AI analyses and individual adjustments
  • Or even "complete feedback solution" as a service provider for customers

Differentiation from other solutions - for the customer

  • Simple user interface
  • Short communication channels to the team so individual adaptations are cost-effective and can be implemented in an extremely short time
  • DSGVO conformity
  • Knowledge in the field of AI for automated text analysis

Special customer segments

The solution is especially interesting for SMEs, as one does not need any training time and has an automatically prepared visual representation of the data that everyone understands.

Business

Benefits

  • Reduced effort through AI: Regarding the evaluation with AI, a team of 4 persons on the university side would probably have had to deal with the data for about a week. By using AI, this was done in a few seconds, so that the university could directly deal with the contents and results of the survey and the data did not have to be laboriously evaluated.
  • Full service: The technical side was taken over by the provider and the survey was transferred into the system. The university was sent a link which it only had to send to the recipients of the survey. Otherwise one person would have had to deal with it for 2-3 days, so time was saved.

The project schedule

  • The university has to submit to the provider the question to be evaluated and the target group that will participate in the survey.
  • The couchsurvey.com team evaluated the technical requirements of the platform together with the university and was able to implement the individual adjustments within one week. During the development of the adaptations, the questionnaire was designed in cooperation with the university.
  • The university sent the survey to the target group and asked them to participate.
  • After the survey had been live for 2 weeks and more than 10,000 answers had been received, the provider analysed and visualised the results, mainly consisting of free text answers, with the help of artificial intelligence, more precisely with procedures from the field of NLP. These results were presented and made available to the university.
  • Regarding the evaluation via AI, a team of 4 persons would probably have had to deal with the data for about one week. By using AI this was done in a few seconds, so that the university could directly deal with the contents and results of the survey and did not have to laboriously evaluate the data itself. The university was extremely satisfied with the results that were presented and would like to implement further projects with the provider in the future. The university has also already explicitly recommended the provider to other universities and educational institutions.

Project maturity level

Used company-wide

Project duration

4 weeks

Project cost
(digits)

4

Running costs
(per month, digits)

2

Involved employees
(Project phase, FTE)

0.1
FTE

Data protection

The customer's data is stored with a cloud provider in Germany and the system used meets all DSGVO guidelines as well as the ISO27001 standard. The customer can log into his personal account on our platform. The site is encrypted using SSL and the account is authenticated via email verification.

Data processed and stored at

EU
US
Switzerland
China
Worldwide

Billing model

One Time Payment
Subscription
Maintenance mandatory

Subscription charged by

Fixed payments
Usage fees

License fee charged by

Accounts
Active user
PC
PC+laptop
CPU
CPU-cores
CPU-threads
Storage
Bandwith
Transfered data

Technical

Skills required by the customer

No technical knowledge necessary.

The project schedule

It is a react frontend with a NoSQL database behind it. Since it is a browser-based solution, there are no technical requirements other than an internet connection.

  • The university has specified the question to be evaluated and the target group to participate in the survey in a video call.
  • The couchsurvey.com team evaluated the technical requirements of the platform together with the university and was able to implement the individual adjustments within one week. The CI of the frontend was adapted to the wishes of the university and other minor changes were made to the system. During the development of the adaptations, the questionnaire was designed together with the university and created on the platform.
  • The university sent the survey to its target group by e-mail and asked them to participate.
  • After the survey had been live for 2 weeks and more than 10.000 answers had been received, the results, which manly consisted of free text answeres were analysed and visualised with the help of artificial intelligence, more precisely with methods from the field of NLP. For this purpose the data was exported from the platform, clustered with the help of SpaCy, an NLP framework, and then prepared graphically for presentation and summarized in a report.

Browser frontend

Headless

Available at mobile

Android App
iOS App
PWA
Browser (almost completely)
Browser (partially)

Deployment

SaaS
SaaS dedicated
Customer server
Private cloud
Workstations
Docker
Helm Chart

Administration

Unnecessary
Commandline
API (completely)
API (partially)
Admin Web UI
Built-in

API (% of features)

0

Integrations available

CSV download of the data

Integration support

Documentation
Phone support
Email support
Implementation service
Implementation by partner
Other

Support

Phone (business hours)
Phone 24/7
Email - response time
Email - best effort
Chat (business hours)
Chat 24/7
FAQ
Knowledge DB
Other

Providers

Christopher Ernst (Projektmanager)

The companies often do not understand that they also face internal costs, e.g. for evaluating the survey results, if they do not want to use a professional and paid feedback solution and it is still more expensive for the company overall than if they would use an external and paid solution.

Christopher Ernst

The information may of course vary in individual cases. Please contact the provider for an assessment of your project.