Détails de l'annonce
Publiée le:
2 février 2021
Type de poste:
Lieu de travail:
El Menzah, Tunis, Tunisie
Entre 1 et 2 ans
Bac + 5
Plein temps
Anglais , Allemand
Secteur: informatique / télécoms
Taille: Entre 100 et 200 employés

Bertsch Innovation GmbH est un leader européen des logiciels d'entreprise standardisés dans le domaine de la gestion des informations sur les produits, la gestion des actifs numériques et à la communication des produits multi-canal.

Dans le cadre de notre expansion en Tunisie , nous cherchons une équipe de développement de logiciels pour notre nouveau local a El Menzah 4.

Description de l'annonce:


Proposal 1: Integrate mediacockpit entity events into apache kafka

mediacockpit has an entity event framework where events related to entities (create, edit, download, delete) are saved to json and in a lucene index for later search and reporting.

mediacockpit entity events can be relevant for other systems, for example, a state change might be important for a shop or ERP system, to indicate an immediate action in that system.

in the future, it could be sensible to integrate mediacockpit with apache kafka for sending and also receiving events. sending would be relevant for internal system events, receiving could be relevant for events e.g. from ERP system (e.g. create new article or change lifecycle flat for article to "end of life") or some usage information from customer-facing system could be streamlined by connecting mediacockpit to an event platform like kafka.

the task of this topic is to implement a generic kafka connector for mediacockpit and integrate both incoming and outgoing events and messages. for the start, the outgoing entity events should be published and (as proof of concept) incoming CREATE or CHANGE events e.g. from an ERP or PLM should be consumed and processed in mediacockpit.

ideally, this functionality would be abstraced as an interface to that the concrete implementation (today with kafka) can be easily replaced by another system in the future.


Proposal 2 : mediacockpit integation in Microsoft Teams

microsoft teams is an increasingly important channel for corporate communication. the collaboration for product and media creation, search and retrieval often occurs in or tangentially to microsoft teams.

to further streamline this process and the integation with teams, mediacockpit should provide a chatbot integration. this integration should respect the existring user permissions (and optionally a "public mode" where users can search view any content marked as "public") and assist customers to search for content via chatbot, then display a visual list of results and optionally share a datasheet (E.g. pdf file). if customer/system supports SSO, then use that SSO context, otherwise provide user a way to login/authenticate (if not already).


at a high level this includes these functions

- authenticate user

- search for / filter content (this is an iterative process based on what is searchable in the system)

- display a result

- transform result into another format (e.g. excel, or pdf proof sheet)

- share this result with others

the process - in particuar the search / filter step(s) should be flexible and support multi-step inquiries, for example:

Hello, I am mediacockpit chatbot, how can I help you?

I am searching for for (media/products)

what media/product would you like to see?

I want pictures of (trains)

OK I found 500 pictures, would you like to see the newest 10 or narrow down your search?

 narrow down search.

OK you can narrow down by keywords or search in specific attributes - just type keywords or say search by attribute with any attribute in this list (dispaly list of searchable attributes)

search in attribute color...


okay I have a list of colors  - dispaly list


that attirbute is free-text, type some terms


that  field is a number, give a range from..to


that is a boolean, do you want to search yes or no...

etc. ideally, the entire search / filter options avaialbe in mediacockpit or our existing plugins can also be searched via chatbot. depending on the data type, the exact method of query and response can and must vary. the most important parameters

- (1) top-level: search product / media

- (1) drill-down by type (asset types or product types)

- (2) filter drill-down via hierarchy / folders

- (1) seach by fulltext/keywords

- (1) search by user defined facets

- (3) search by other attributes / data types

The results of queries should be listable as image thumbnails in the chat or as a list of keys, optionally downloadable as an excel/csv file...... the additional metadata highlights (to be defined) can also be displayed or included in the download file (to create this file, the chatbot could leverage existing system functionality for publishing "e.g. DCM export or webDCM".

in the case of media, the option to download the original file (or a set of one or more renditions) directly should be possible.

in the best case, a result should be sharable with others in microsoft teams. in that case, the "external selection" functionality - that is, the permissions of the initiator are valid and the recipient does not need additional authenticaiton..

open issues-  how to handle chatbot in teams rather than 1:1 chats, where permissions cannot be drilled down to one single user 

Proposal 3



We are looking to build a best practice mc cluster configuration that can scale


  • scale with data (asset store e.g. using s3)
  • scale for load (new cluster nodes)

and also easy to centrally manage and monitor


  • central logging
  • central monitoring