One day, hands-on. You don’t leave with impressions of a software factory — you leave with one running.
01 — WHAT MAKES IT SPECIAL
Most teams have tried coding agents. Few deliver software with them. The gap isn’t the model — it’s the factory around it: orchestration, context, validation, governance.
This bootcamp is not an intro to AI coding tools. It’s one day of building the system your team uses to ship — and you run it before you go home.
01
A plan-first pipeline — run twice
Working agent setup with planning, minimal orchestration and sub-agents. You run it once on a greenfield app, once on real work from your backlog.
02
Your factory map
Your own SDLC, rules and processes mapped against the six capabilities every software factory needs: orchestration, isolation, context, validation, governance, learning.
03
Encoded artifacts
A context file with your team’s conventions, at least one validation gate, and skills published to a shared skill marketplace — built by you, taken home.
04
Real work delivered agent-first
A task from your own backlog in your own codebase — or an extension of the morning’s greenfield app with your new skills.
05
Whole product teams, one table
50 participants — developers, product engineers, PMs. Non-technical participants run the same exercises with adapted scope. Tables that keep talking after the event.
06
Full cost transparency
€999 per person, all in — sessions, hands-on work, meals, drinks. Every additional ticket €799. No hidden add-ons. No sponsors.
02 · the format
Every loop follows the same cadence: short input, hands-on practice, reflection. No big blocks of theory, no unbroken lab time. You build with your own agent setup, at a facilitated table — and close the day with a marketplace showcase of skills tested on real code.
22.
SEPT
2026
ON
BER
BERLIN
Delta Campus
ON
50
MAX
Participants
ON
4
LOOPS
Hands-on
ON
02.1 · factory capabilities
Four of them you build hands-on during the day. Isolation and governance at enterprise scale are framed in the keynote and land on your roadmap — honestly marked below.
HANDS-ON
Context
Your team’s conventions encoded in context files agents actually follow. You write yours in Loop 3.
ROADMAP
Isolation
Safe spaces for agents to work in parallel without breaking each other — framed in the keynote, on your roadmap.
ROADMAP
Governance
Permissions, budgets and accountability at enterprise scale — framed in the keynote, on your roadmap.
HANDS-ON
Validation
Gates that catch bad output before it ships: rules, review gates, tests. You build at least one during the day.
HANDS-ON
Learning
Skills that improve with every run — published to a shared marketplace and adopted across tables.
HANDS-ON
Orchestration
Plan first, then execute — with sub-agents where they pay off. The minimal pipeline you run twice during the day.
02.2 · impressions
03 · TRAINERS
People who've done it. Not people who only talk about it.
Stefan Munz
//Trainer & Curriculum Lead
Hosted the first agentic coding meetup in Hamburg in July 2025 (probably the first in Germany), and co-founded hackers & wizards with Benedikt in 2026. Now full-time on helping teams make AI coding agents actually work in production.
Benedikt Stemmildt
//Curriculum Lead
20+ years in enterprise engineering at OTTO & BLUME2000. Founder of hackers & wizards. Now 100% focused on helping teams turn agentic engineering into systematic practice.
Tereza Iofciu
//AI & Agentic Engineering Coach
PhD in Computer Science. 15+ years leading data and AI teams at neuefische, FREE NOW, and New Work (XING). Now coaching data leaders on navigating the shift to AI-driven engineering.
Your trainers for the day.
04 · Agenda
// KICKOFF
09:00 – 10:20
09:00
Arrival, Setup Check, Coffee & Snacks
Informal arrival, first conversations, verifying your agent setup runs. No content yet.
09:30
Opening Keynote: From AI Experiments to the AI Factory
Not an intro to AI coding tools, but how teams deliver software with agents. Introduces the six factory capabilities — the frame for the whole day.
10:00
Introduction Round: AI Reality Check
Share your current maturity: what you already tried, what worked, what failed, where you are blocked.
// LOOP_1 · GREENFIELD
10:20 – 12:00
10:20
Hands-On: Greenfield, Agent-First
Everyone builds the same application with a coding agent: plan first, minimal orchestration, sub-agents where they pay off. You run the factory pipeline before lunch.
// LOOP_2 · MAP YOUR FACTORY
12:00 – 13:35
12:00
Map Your Factory
Your SDLC, rules and processes mapped against the six factory capabilities: what exists, what’s missing, where it lives. This map drives the rest of your day.
12:35
Lunch Break
Time to breathe, eat and compare factory maps across tables.
// LOOP_3 · ENCODE & MARKETPLACE
13:35 – 15:15
13:35
Hands-On: Build the Skill Marketplace
Turn your team’s conventions into context files, validation gates and skills. Publish them to a shared marketplace, and take what other tables built.
// LOOP_4 · BROWNFIELD
15:15 – 16:45
15:15
Hands-On: Brownfield, Your Own Codebase
Real tasks from your backlog, agent-first, in your own repo — where the factory meets legacy code. No use case with you? Extend the morning’s greenfield app with your new skills.
// SHOWCASE & CLOSE
16:45 – 18:00
16:45
Marketplace Showcase & Awards
Every table demos the skills that survived contact with real code. Before/after against the morning build. Awards for best skill, best validation gate, best steering.
17:45
Commit & Close
Your factory roadmap: the next capability you build after today. Group photo.
05 · MARKETPLACE
Skills that survive real code.
Marketplace showcase & awards at 16:45.
During the day, every table publishes skills, context files and validation gates to a shared skill marketplace — and adopts what other tables built.
The closing showcase is the proof: which skills held up against real code, before/after versus the morning build.
Awards for best skill, best validation gate, and best steering.
06 · OUTCOME
Not impressions — a working factory core, a small unit of production-ready work delivered with it, and a clear plan for what to build next. Here’s what goes home with you:
Agent Setup & Plan-First Pipeline
A working agent setup you have run twice: once on a greenfield app, once on real work.
Your Factory Map
Your SDLC, rules and processes mapped against the six capabilities — with the gaps marked.
Encoded Artifacts
A context file with your team’s conventions, at least one validation gate, and skills in the shared marketplace.
Real Work, Delivered Agent-First
A task from your own backlog in your own codebase — or the extended greenfield app.
Extension Roadmap
Which capabilities your factory already has, which are missing, and which one to build next.
07 · target
Engineering Manager
Senior Developer
Software Architect
Product Manager
Tech Lead
Agentic Engineer
Product Engineer
Head of Engineering
// 22.9.2026 · Hamburg
06 · pricing
SINGLE_TICKET · +799 € for every additional ticket
All sessions, keynotes, live demos and hands-on workshops
5 pods with expert trainers
Full starter kit: CLAUDE.md, rules, skills, MCP configs
90-day rollout template, built during Day 2
All meals & drinks across both days
alphalist community access and premium networking
SEATS_LEFT: 46
Save your seat
// questions? → cto@alphalist.com
Powered by alphalist — Germanys no. 1 CTO Community and organizer of the code.talks developer conference.
© 2026 alphalist GmbH