Azure Developer Community Day 2022

December 13, Microsoft Germany, München

Register now!

Alexander Wachtel

Microsoft AI MVP, MCT, CEO @ ESC Deutschland & Ph.D. Candidate @ Karlsruhe Institute of Technology

Alexander Wachtel

Alexander Wachtel is Microsoft AI MVP and is working as a Managing Director at ESC Germany. As a Microsoft Gold Partner in several competencies, ESC Germany offers professional software engineering, consulting, and training services for Microsoft products and technologies. Due to his many years of experience, he has been honored as a Microsoft Certified Trainer (MCT) since 2013 till today.

After graduating in computer science at the Karlsruhe Institute of Technology (KIT) in 2012, he also worked on his Ph.D. thesis in the field of programming in natural language at the chair of Prof. Dr. Walter F. Tichy at KIT. During this time, he has published several technical research articles in the area of Natural Language Processing, Natural Language Programming, AI & Cognitive Services.

Furthermore, he has integrated Microsoft technologies into his KIT courses, and still teaching software engineering at the KIT using Microsoft technologies, especially Azure DevOps. For this reason, due to his teaching activities at the KIT, he was also awarded as a Microsoft Innovative Educator (MIE) expert from 2016 to 2019.


Sessions

Github Copilot: Ihr KI Assistent beim Programmieren @ 13:30

Der Vortrag behandelt den GitHub Copilot von OpenAI. Aufbauend auf einem Encoder-Decoder Modell wird durch den GPT-3 Transformer ein Werkzeug für die Sprachverarbeitung implementiert. Dabei erstellt der GPT-3 Transformer mithilfe eines Aufmerksamkeitsmechanismus natürlichsprachliche Texte, die von denen eines Menschen kaum zu unterscheiden sind. Zusammen mit dem Model Codex, das von Guthub und OpenAI gebaut ist, ergeben sich dadurch zahlreiche Möglichkeiten, zu einer natürlichsprachlichen Eingabe des Entwicklers passenden Programmiersnippets vorzuschlagen. GitHub Copilot ist jedoch mehr als die automatische Vervollständigung erkannter Sprachmuster. Es verwendet den bereitgestellten Kontext, um einen passenden Code zu synthetisieren.