docs: add kickoff meeting notes (28.04.2026) (#2)

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot <copilot@github.com>

Author: 3 people

docs/meeting_notes/2026-04-28-kickoff.md

@@ -0,0 +1,82 @@
+# Kickoff Meeting
+
+**2026-04-28, 16:00 – 18:30**
+**Participants:** Manuel, Florian, Leon
+
+---
+
+## Work Division
+
+No hard splits decided yet. The plan is to work wherever work is needed and pick up tasks based on what makes sense at the time. We want to avoid people being blocked because "that's not my part."
+
+Everyone keeps an eye on infra/CI/CD, no one owns it alone.
+
+## Rough ownership allocation
+
+Below is a suggested, non-exhaustive allocation of subsystem leads. This is a rough guideline and will be finalized with the tutor; anyone may still contribute to other subsystems.
+
+- **Incident Service (backend):** Florian (lead)
+- **Event Log & Database:** Florian (lead)
+- **GenAI service & prompts:** Manuel (lead)
+- **Frontend / Demo UI:** Leon (lead)
+- **Rule Engine / Automation:** Florian (lead)
+- **Infrastructure, CI/CD & Monitoring (Prometheus/Grafana):** Shared (all) — primary contact: Manuel; all team required for reviews
+- **Testing, QA & E2E:** Shared (all)
+- **Docs & Meeting notes:** Manuel (lead)
+
+---
+
+## Finding a Project Topic
+
+Went through a list of proposals we each came up with beforehand:
+
+- ~~Proposal 1 — AI Study Assistant~~ → weak system scope, basically a chatbot wrapper
+- ~~Proposal 2 — AI Knowledge Base (RAG System)~~ → interesting technically, but no clear application domain / user story
+- **Proposal 3 — AI DevOps Incident Assistant** → picked this one (Leon's idea — he encountered this kind of tooling during a deployment at a company)
+- ~~Proposal 4 — AI Meal Planner & Grocery Optimizer~~ → GenAI angle fine, but domain felt arbitrary
+- ~~Proposal 5 — AI Travel Planner~~ → same issue, doesn't connect to the course theme
+
+Landed on **Proposal 3** fairly quickly once we talked through it:
+
+- fits the team name (panic at the console)
+- real GenAI use case: summarize incident context, suggest severity, draft postmortems
+- event-driven architecture is natural here → good fit for microservices
+- connects directly to the DevOps theme of the course
+- CI webhook integration makes for a clean demo story (pipeline fails → incident auto-created)
+
+Manuel sketched a rough service breakdown using ungodly amounts of LLM calls, Florian wrote things down and helped break the system into concrete subtasks and microservices, Leon started thinking about what the AI service would actually do.
+
+Main concern raised: is this too complex for 3 people? → Agreed to keep the core small (incident service + event log + frontend) and treat the rest as extensions.
+
+**Key principle we agreed on:** GenAI must be a real feature, not decoration. It reads the event log and incident metadata and produces something useful (summary, triage, postmortem draft). Not just "ask ChatGPT".
+
+---
+
+## Rough Architecture Idea
+
+Nothing committed yet, but the rough idea on the whiteboard:
+
+- Frontend talks to a gateway / backend
+- Backend: at least Incident Service, Event Log Service, maybe a Rule Engine for auto-creating incidents
+- GenAI: separate Python service, stateless, reads incident context, returns structured JSON
+- DB: PostgreSQL
+
+Detailed architecture → next meeting
+
+---
+
+## Deferred to Next Meeting
+
+- Finalize microservice breakdown
+- Folder/repo structure
+- Backlog + user stories
+- UML diagrams (use case, component, object model)
+- Decide on message broker (Kafka? just REST for MVP?)
+
+---
+
+## Next Steps
+
+- Find tutor meeting slot
+- Ask tutor about how he thinks the meetings should go. What to expect from us, how to prepare, etc.
+- Finalize ownership allocation with tutor

docs/meeting_notes/2026-04-28-task_distribution.md

@@ -0,0 +1,34 @@
+# 👥 Task Distribution
+
+This document defines team responsibilities for the initial deliverable. Subsystem ownership will be decided later.
+
+---
+
+## Initial Deliverable (Problem Statement + System Overview — due 08.05.2026)
+
+### Manuel
+
+- Problem Statement
+- High-level project description
+- Definition of main functionality
+- Repo structure setup
+
+### Florian
+
+- First Product Backlog
+- User stories
+- Feature priorities
+
+### Leon
+
+- UML Diagrams:
+  - Analysis Object Model (UML Class Diagram)
+  - Use Case Diagram
+  - UML Component Diagram (Top-level architecture)
+- Write tutor (contact + coordinate meeting)
+
+---
+
+## Collaboration Note
+
+No hard subsystem splits for now. Everyone picks up work where it's needed. All members collaborate on integration, CI/CD, deployment, and code reviews.