Testing code in an adversarial setting: a case study

We are happy to announce our next meetup that will be held physically at XM, in Athens.
After the talk there will be time for Q&A, food and networking.

📍 Location 📍
Michalakopoulou 80 street, 1st floor.
Attendees should register with their meetup name at the entrance

⏰ Date & Time ⏰
24th of September, 19:00

🔗 Prerequisites & Target audience 🔗
The talk is aimed at developers who are interested in C++.

🇬🇷 Language 🇬🇧
The talk will be in Greek unless otherwise requested.

🍀 JetBrains license ruffle 🍀
At the end of the event we will ruffle out 2 yearly licenses for a JetBrains IDE.

📃 Agenda 📃
19:00 - Intro, Agenda of meetup
19:05 - A message from our sponsor - XM
19:10 - "Testing code in an adversarial setting: a case study" by Vaptistis Anogeianakis
20:10 - Pizza, kindly offered by XM 🍕

🎤 About the talk 🎤
We will dig into a deceptively simple problem: implementing a function to compute N choose K (binomial coefficients). The real challenge isn't the math, it's handling arithmetic overflow correctly.
We will walk through a related problem posted on CodeSignal, the optimal solution the speaker had in mind, the range of solutions submitted by participants, and the custom C++ framework built to evaluate them.

The framework is basically a poor-man's version of the back-end infrastructure powering sites like LeetCode, CodeSignal, and HackerRank. It was built because standard unit and integration testing tools do not address the specific problems that come up when evaluating this kind of solution at scale.

Finally, the speaker will discuss an alternative way of evaluating engineers, in exams, interviews, and similar settings, and more recently, evaluating LLMs. It wraps up with observations from trying this method out on a group of young programmers.

🎤 About the speaker 🎤
Vaptistis Anogeianakis is a senior software engineer and junior inventor.
For the last 6 years or so, he was leading the collection and analysis of telemetry data at the product level at IBM.
Before that, he was designing virtualization management algorithms in Java, crowd-funding the development of his own language, and had spent 10 years developing open source projects in C++.

His interests include computer graphics and programming languages, and he has a patent pending in the virtualization ecosystem.