When your feed is flooded with event announcements full of vague slogans about "the future is already here," the first thing I check is the speakers and the agenda. This November, I came across two free engineering events where, instead of abstract theory, speakers break down concrete architectural trade-offs: defending neural-network pipelines from attacks and bringing AI agents into team development with strict token and access controls. I pulled together the verified details for both events so you can plan your calendar ahead.
Open Online Lectures by Центральный университет: AI System Security and Product Skills
- Organizer: Центральный университет (Школа технологий).
- Date and time: November 17 and 18, 2026, broadcasts start at 19:00 Moscow time (MSK, UTC+3).
- Format and terms: online livestream, free attendance, advance registration required on the program page.
- Audience: developers, information security specialists, product managers, and data analysts.
The lecture series focuses on practical shifts driven by broad adoption of neural networks in a company's engineering stack:
- November 17 (19:00 MSK) — «Как злоумышленники атакуют современные ИИ-системы»: Georgy Kucherin, Senior Threat Researcher (GReAT) at Лаборатория Касперского, will dissect vulnerability exploitation mechanisms in neural-network pipelines, prompt-injection techniques, bypassing built-in system guardrails, and practical methods for model isolation.
- November 18 (19:00 MSK) — «Навыки продуктовых менеджеров в эпоху ИИ»: Alexander Kapustin, Product Director of the Auto segment at Т-Банк, will show how collaboration between product managers and engineering teams is changing, how to write requirements for ML-adjacent projects, and how to evaluate the unit economics of hypotheses.
Registration is available via the direct link: register for the Центральный университет open lectures.
DevOps Lounge Meetup: AI in Production and in DevOps Engineers' Workflows
- Organizers: Orion soft and the IT conference team Онтико.
- Date and time: November 26, 2026, 18:30–22:00 Moscow time (MSK, UTC+3).
- Format and terms: hybrid (in-person in Moscow with networking plus a live online stream for remote attendees). Free attendance, advance registration required (confirmation with the venue or stream link is emailed by the organizers).
- Audience: DevOps and MLOps engineers, systems architects, and infrastructure platform tech leads.
The agenda features two deep engineering talks grounded in real production experience:
- «AI Gateway в Nova AI: одного эндпоинта мало» (Dmitry Matushkin, MLOps Engineer at Orion soft): the talk covers enterprise model-access infrastructure. Dmitry will explain how to centralize traffic through an AI Gateway, implement role-based access control (RBAC), configure token quotas and limits across services, and set up tracing, smart load balancing, and automatic failover to backup models during outages. Special focus goes to integrating guardrails to prevent personal data (PII) leaks and filter unwanted output.
- «Опыт работы с коллегой-агентом: от локального харнесса к мультиплееру с агентами в командных тулах» (Sergey Kharlamov, Head of Product Lab at VK Tech): a walkthrough of moving from isolated coding-assistant experiments to full-fledged team collaboration. Sergey will demonstrate the architecture of a virtual teammate wired into corporate webhooks that listens to Jira and chat events, handles routine documentation prep and pull-request reviews, and rolls out changes to staging strictly after team-lead approval.
After the main talks, an interactive game and informal networking are planned for in-person attendees. You can register for the meetup or join the livestream at: register for DevOps Lounge.

![Node.JS [ru]](/api/digests/it_development/daily/20261112/assets/sources/we-use-js.jpg?v=96d603ea907a603e2476230869b24b042f87006317785f577b52227c3db5cc47)