Blätter-Navigation

Of­f­re 73 sur 116 du 08/01/2020, 16:45

logo

Tech­ni­sche Uni­ver­sität Ber­lin - Fac­ulty IV - Insti­tute of Tele­com­mu­nic­a­tion Sys­tems / Com­plex and Dis­trib­uted IT Sys­tems (CIT)

2 pos­i­tions - Research Assist­ant - salary grade E13 TV-L Ber­liner Hoch­schu­len

under the reserve that funds are gran­ted - part-time employ­ment may be pos­sible

Design­ing a reli­able sys­tem for Cloud Main­ten­ance and ensur­ing resi­li­ence (anom­aly detec­tion, root cause ana­lysis and recov­ery) is chal­len­ging due to the com­plex nature of the sys­tem. Resi­li­ence is defined as the abil­ity of a cloud plat­form to recover quickly and con­tinue oper­at­ing even when there has been a fail­ure, also referred to as fault tol­er­ance. Efforts are made to imple­ment pre­dict­ive approaches that can remedi­ate sys­tem parts exhib­it­ing anom­alies, which not yet evolved to sys­tem fail­ures. Recently, AIOps evolved as a cat­egory aim­ing at the com­bin­a­tion of sys­tem oper­a­tion with arti­fi­cial intel­li­gence (AI) meth­ods. Thereby IT oper­a­tions are auto­mated and enhanced by using machine learn­ing and AI to ana­lyze data streams col­lec­ted from vari­ous IT mon­it­or­ing tools and devices. This allows to react to issues in real time.

Work­ing field:

The aim of the pro­ject is to do research and develop AIOps meth­ods for con­tinu­ous data streams. One chal­lenge is to cre­ate solu­tions, that are gen­eral enough to oper­ate across dif­fer­ent sys­tems. We focus on the fol­low­ing top­ics: log and met­ric data embed­ding, learn­ing joint rep­res­ent­a­tions for mul­tiple sys­tems, anom­aly detec­tion, and explain­ing anom­alies. All these will entail design­ing a gen­eral method, imple­ment­ing a pro­to­type in the con­text of exist­ing open source pro­jects, and exper­i­ment­ally eval­u­at­ing the pro­to­type with a test data using sim­u­lated and pro­duc­tion data.

Require­ments:

  • Suc­cess­fully com­pleted uni­versity degree (Mas­ter, Dip­lom or equi­val­ent) in Com­puter Sci­ence, with spe­cial­iz­a­tion in dis­trib­uted sys­tems. Com­pleted courses in arti­fi­cial intel­li­gence and data sci­ence. Prefer­ably exper­i­ence in the field of AIOps.
  • Interest in sys­tem devel­op­ment and oper­a­tion of large-scale soft­ware archi­tec­ture, as well as enthu­si­asm to estab­lish recent research res­ults in prac­tice.

Min­imum qual­i­fic­a­tions:
  • Exper­i­ence with stat­ist­ical soft­ware (python, numpy);
  • Exper­i­ence with data­base lan­guages (SQL, mon­goDB);
  • Exper­i­ence with big data plat­forms (Hadoop, SPARK);
  • Exper­i­ence in work­ing with cluster and cloud sys­tems (Open­Stack);
  • Build­ing and oper­a­tion of con­tain­ers (Docker).
Pre­ferred qual­i­fic­a­tions:
  • Prac­tical exper­i­ence in pro­ject man­age­ment and agile devel­op­ment meth­od­o­lo­gies;
  • Prac­tical exper­i­ence in con­cep­tion and design of AI sys­tem solu­tions;
  • Famil­iar in work­ing with meth­ods from the domain of real time and stream data ana­lysis;
  • Exper­i­ence and interest in the top­ics of machine learn­ing, rep­res­ent­a­tional learn­ing, anom­aly detec­tion and time series clas­si­fic­a­tion;
  • Exper­i­ence in work­ing with explain­able machine learn­ing meth­od­o­lo­gies;
  • Exper­i­ence with Tensor­Flow/PyT­orch/keras;
  • Exper­i­ence in writ­ing of sci­entific papers.

Fur­ther require­ments include team spirit and an excel­lent com­mand of the Ger­man and Eng­lish lan­guages (spoken / writ­ten).

How to ap­ply:

Please send your writ­ten applic­a­tion with the ref­er­ence num­ber and the usual doc­u­ments to Tech­nis­che Uni­versität Ber­lin - Der Präsid­ent - Fakultät IV, Institut für Telekommunikationssysteme, FG Komplexe und Verteilte IT-Systeme, Prof. Dr. Odej Kao, Sekr. TEL 12-5, Ernst-Reuter-Platz 7, 10587 Berlin or by e-mail to odej.kao@tu-berlin.de.

To ensure equal opportunities between women and men, applications by women with the required qualifications are explicitly desired. Qualified individuals with disabilities will be favored. The TU Berlin values the diversity of its members and is committed to the goals of equal opportunities.

Please send copies only. Original documents will not be returned.