Claude Community Wrocław 4 [EN]

[Event in English 🇬🇧]
Welcome to the third Claude Community Wrocław meetup - a community-driven initiative built by builders, for builders https://claudecommunity.pl/

We share real stories of how we use Claude, Claude Code, and Claude Cowork in our daily work. No product pitches, no vendor talk - just honest experience from people who build with these tools every day.

This event is NOT affiliated with Anthropic. This is an independent community initiative.

Pizza and drinks are included and the event is free, sponsored by CSHARK
If you wish to speak at our future events, use this form: https://forms.gle/8fRNxrCG8JWJUkRQ9

🟢 Where and when?
Concordia Design Wrocław (CSHARK office), wyspa Słodowa 7, 50-266 Wrocław
6PM CEST 🔴 the location is different than usual 🔴

Agenda
➡️ 18:00 - Welcome & intro by Claude Community Wrocław organizers
🧑‍💻 18:05 - Trust, but eval: from one skill to a team of agents and proving each step actually works - Konrad Dżbik
🧑‍💻 18:45 - What’s the actual ROI of AI in your SDLC? Measuring it with the metrics we already have - Maks Piechota
🍻 19:25 - Networking & Pizza
🧑‍💻 19:55 - Lowering the Floor, Not the Ceiling: Bringing Non-Technical Users into Agentic AI - based on real example - Marcin Krok

➡️ Trust, but eval: from one skill to a team of agents - and proving each step actually works

Everyone ships skills, plugins and multi-agent setups for Claude Code. Almost nobody can prove they help – or say what they cost. I'll walk up the autonomy ladder with real artefacts from a 200-person software house and show how the same discipline – a baseline, a grader and a budget – applies at every rung.

🎤 Konrad Dżbik - I'm Head of AI at CSHARK\, a software house in Wrocław\, where I built the AI Vertical from a blank page and brought Claude into everyday work across the whole company. I run Open mAIc\, a workshop series that gets developers\, testers\, PMs\, marketing\, and finance learning in the same room\, and I keep teams supplied with skills and plugins through our in-house CShark-Hub marketplace. I started out writing C/.NET\, hold a Google Cloud Professional Cloud Architect certification\, and I'm a certified Bielik trainer with a soft spot for Poland's sovereign LLM scene. My natural habitat is Claude Code\, agentic systems\, MCP\, and multi-agent orchestration — and my firm belief is that an agent you haven't evaluated is just a very confident guess.

➡️ What’s the actual ROI of AI in your SDLC? Measuring it with the metrics we already have

A lot has already gone into AI tools: licences, tooling and a good deal of your teams’ time. So what’s the return? There’s no clean formula for it yet, but measuring delivery performance isn’t a new problem. My thesis is that AI in the SDLC is one brick, not a new building. What you’re after is still engineering performance, so the metric set stays broadly the one we already had, with some tactical extensions for what’s genuinely specific to agents. We’ll look at the data your AI vendor is already collecting on your behalf, DORA and what its authors say about measuring AI-assisted delivery, DX Core 4 and its tension metrics read against each other so no single number can be gamed on its own, the Lean Startup learning loop for connecting engineering work to business outcomes, and the Product Operating Model.

🎤 Maks Piechota - I’m a Senior Engineer on the AI Platform team at FirstConnect Insurance Services. My whole career\, from enterprise (Dolby Laboratories) to US startups\, has centered on one thing: making software delivery faster without giving up quality\, through demanding Engineering Excellence practices (TDD\, BDD\, Extreme Programming\, Trunk Based Development\, Continuous Delivery) backed by hands-on software architecture. As Head of Engineering at Bridge I went deep into DORA and DX\, and that became the baseline I brought to leading AI adoption into the SDLC at an American insurtech. I believe the practices the industry spent decades building to minimize human-error risk are exactly what AI-generated code needs\, and I share what I learn at meetups and conferences.

➡️ Lowering the Floor, Not the Ceiling: Bringing Non-Technical Users into Agentic AI — based on real example

Agentic AI keeps raising its ceiling while the floor to get in stays high — install a toolchain, live in a terminal, wire up credentials. Through a real product built on Claude Code — one-button dependency install, shared knowledge bases, agents in isolated git worktrees under budget caps and merge review — and real onboarding stories, I'll show where the floor actually sits for non-technical users and the engineering it takes to lower it without dumbing anything down. The last mile of AI adoption is rarely the model — it's the setup screen, and it's ours to fix.

🎤 Marcin Krok - I'm an AI enthusiast and co-founder of CSHARK, a Wrocław-based software house I've been building since 2014. A former software developer with 25 years of experience in IT — including 15 years running multiple businesses in the software space — I spend a lot of my time these days wrestling with what it actually takes to get AI working inside a real engineering org, not just in a demo. That's also what's behind the examples I'll be sharing at this talk: real onboarding stories from a Claude Code-based tool built to get non-technical teams past the setup screen and actually shipping with AI agents.

Sponsor

CSHARK (https://www.cshark.com/)