TALENT_ARCHITECTURE_v2026 // POSITIONS_OPEN

BUILD THE
FUTURE.
FROM CODE.

// We don't hire for roles. We hire for convictions. IRONCODE is assembling its founding team — the people who will architect digital sovereignty for Europe's enterprises and re-imagine the organizations that run on it.

4 POSITIONS OPEN
THREE
FILTERS.

Every person who joins IRONCODE must clear three thresholds. Not skills tests. Not culture-fit theater. Convictions.

01
AUTONOMY

Can you walk into a client's CTO office on Monday and lead the room? IRONCODE has no project managers. No one tells you what to do next. You architect the outcome and own it completely.

02
RANGE

Can you architect AND build? Advise AND ship? Strategy AND code? Specialists belong at the big four. IRONCODE needs people who operate across the full stack — technical, organizational, and commercial.

03
VELOCITY

Do you move at the speed of the work or the speed of the process? We ship week one — not a plan, an artifact with code in it. Comfort with ambiguity isn't optional. It's the baseline.

04
CONVICTION

Do you believe the enterprise consulting market is broken? That most "digital transformation" is analog preservation with a cloud subscription? That AI doesn't just automate work — it re-imagines what work is? Good. We wrote the manifesto for you.

People who need to be told what to do
People who write slide decks instead of code
People who think "agile" means "more meetings"
People who optimize for comfort over impact
People who've never deployed something real users depend on
WHAT FOUNDING
LOOKS LIKE.

This isn't a job. It's a founding position. The difference matters — in how you work, what you own, and where this goes.

EQUITY //
OWNERSHIP

Founding team members participate in IRONCODE's equity program. You're building this — you should own a piece of what you build. Vesting structure designed for long-term alignment, not golden handcuffs.

AUTONOMY //
ZERO SURVEILLANCE

Remote-first within European timezone. No tracking, no time sheets, no utilization reports. Outcome-measured. You own the result — how you get there is your architecture decision.

TOOLS //
HARDWARE + LEARNING

MacBook Pro M-series. Annual learning budget (conferences, certifications, courses). You'll also get access to IRONCODE's full AI toolchain — Claude, GitHub Copilot, and every agent we build internally.

IMPACT //
REAL CLIENTS

Week one, you're in a CTO conversation. Not in orientation. Not watching onboarding videos. We work with mid-market and enterprise clients on real systems with real stakes.

SPEED //
SHIP EVERY WEEK

No six-month roadmaps. No sprint theater. No "waiting for approvals." We architect, build, and deploy. Every engagement produces working code, running systems, or re-imagined organizations — not documentation.

FRONTIER //
BUILD BEFORE OTHERS

You'll work on IaaS — our proprietary AI interviewing platform — and deploy multi-agent systems for enterprise clients. This is the frontier of what's possible with agentic AI. You'll be there before most people know it exists.

4 ROLES.
FOUNDING LEVEL.

Berlin and Bucharest. All founding team positions. All require conviction, range, and the ability to lead from day one.

ROLE_02 //
LEAD AI ENGINEER
Berlin / Remote EU OPEN — CRITICAL
+
"Build the AI engine that powers everything IRONCODE delivers. You architect and ship the IaaS platform, design multi-agent systems for clients, and set the technical standard for every AI engagement. You don't demo chatbots — you deploy autonomous agent fleets."
Production LLM applications — not prototypes, not notebooks. Systems that handle real users at scale.
Multi-agent or agentic AI systems in production, using frameworks like LangGraph, CrewAI, Autogen, or Azure AI Foundry
RAG architectures with vector databases (pgvector, Pinecone, Weaviate, Qdrant) — including hybrid retrieval and re-ranking
Production Python (FastAPI, async patterns) — not academic Python
AI systems with 1,000+ concurrent users. You understand the difference between a demo and infrastructure.
LLM orchestration at scale: prompt engineering, structured outputs, function calling, streaming, multi-provider strategy
Agentic system design: orchestrator-task patterns, state management, error recovery, human-in-the-loop patterns
AI governance and safety — EU AI Act awareness, constitutional AI approaches, audit trails
Event-driven architecture. CQRS. Distributed systems trade-offs. Technology choices that survive 3 years.
Voice AI pipeline (Deepgram, ElevenLabs, WebRTC) is a significant plus for our IaaS platform
5-8 years in software engineering, 2+ years focused on LLM/AI systems in production
You write technical content or speak at meetups — you have opinions worth sharing
Comfortable presenting to non-technical stakeholders. Architecture without communication is architecture nobody uses.
Self-organizing — you don't wait for a product manager to tell you what to build
ROLE_03 //
SENIOR PLATFORM ENGINEER
Bucharest, Romania OPEN — CRITICAL
+
"Own the infrastructure that everything runs on. Kubernetes clusters, CI/CD pipelines, IaC, observability — and the IaaS platform infrastructure that must scale to 2,500 concurrent AI interviews. You build the foundation that makes our sovereignty narrative real: bare-metal, on-prem, hybrid cloud. Not just 'we use AWS.'"
Production Kubernetes clusters — not minikube, not managed click-ops. Deep operational experience.
Terraform or Pulumi modules you wrote and own, not copied. You understand the provider model.
GitOps workflows (ArgoCD or Flux) in production with real environment promotion
CI/CD pipelines: GitHub Actions or GitLab CI, blue/green and canary deployment strategies
Infrastructure for applications serving 1,000+ concurrent users under real load
Kubernetes internals: pod security, network policies, HPA/VPA, service mesh (Istio), multi-tenant clusters
Observability: OpenTelemetry, Prometheus, Grafana, distributed tracing, log aggregation
Security: zero-trust networking, Vault for secrets management, compliance-as-code
Bare-metal and hybrid cloud — you're not a cloud-only engineer. You're comfortable with the full spectrum from managed services to on-prem AI inference (vLLM, Ollama).
EU data residency requirements and how to implement them architecturally. Azure AKS preferred (IaaS runs on Azure).
8+ years in infrastructure / platform / DevOps engineering
CKA or CKAD certified (or equivalent demonstrable depth)
You've been on-call. You've handled production incidents. You write your own runbooks.
Comfortable working asynchronously with a Berlin-based team — autonomous by default, not by instruction
ROLE_04 //
SENIOR FULL-STACK ENGINEER
Bucharest, Romania OPEN — CRITICAL
+
"Build the IaaS platform — frontend to backend. The real-time interview interface, the analytics dashboards, the multi-agent orchestration layer, the export pipeline. You turn a concept document into working software that 2,500 people can use simultaneously to have AI-conducted conversations."
A real SaaS product — not internal tools, not agency websites. Software that paying customers depend on.
Real-time applications: chat, collaboration, streaming, live dashboards. You understand WebSocket architecture.
FastAPI backend with async patterns, dependency injection, background tasks, WebSocket handling
Next.js frontend: App Router, server components, real-time data patterns, responsive design
PostgreSQL at depth: schema design, query optimization, partitioning, pgvector for semantic search
Multi-tenant SaaS architecture: org-level isolation, RBAC, SSO integration
LLM API integration (Claude, GPT or equivalent) in a production application
Testing as a default, not an afterthought: pytest, React Testing Library, integration testing
UX opinions — you don't just implement designs, you question them. Product instinct matters here.
Voice/WebRTC experience (Deepgram STT, ElevenLabs TTS) is a significant plus for our IaaS platform
6+ years in full-stack software engineering — genuine full-stack, not frontend-with-some-Python
You can take a concept document and turn it into working software without 50-page specs
Side projects or open-source work showing product-building instinct, not just feature execution
Comfortable working asynchronously with a Berlin-based team
ROLE_05 //
BUSINESS & OPERATIONS LEAD
Berlin, Germany OPEN — CRITICAL
+
"Run everything that isn't architecture or code. Proposals, contracts, finance, legal coordination, partnerships, client operations, and the business machine that lets the technical team focus on technical work. You're the operational backbone of IRONCODE — the one who makes sure the GmbH runs, the invoices go out, and the partnerships get signed."
Operated inside a technology consultancy, IT services firm, or B2B SaaS company — not consumer, not e-commerce
Written proposals and SOWs for consulting engagements. You know the standard terms and where to push.
Managed P&L, cash flow, and forecasting for a growing business — not just tracked it, owned it
Coordinated legal: contracts, NDAs, data processing agreements, employment agreements
Navigated German business infrastructure: GmbH operations, DATEV, Handelsregister, Steuerberater coordination
Value-based pricing — you can defend it in a client negotiation, not just explain it in a slide
German public funding landscape (Gründungszuschuss, ZIM, Forschungszulage, EU grants) is a strong plus
Hyperscaler partner programs (AWS, Azure, Anthropic) and enterprise certification processes
AI tools — not awareness, active daily use. You work with AI-generated proposal drafts and improve them, not resist them.
Multi-entity / cross-border operations (DE + RO or similar) is a strong plus
5-8 years in business operations, consulting operations, or startup operations
Founder energy — you treat the business as your own, not someone else's company to manage
Client-facing comfortable: commercial conversations, not just back-office
You create systems, not processes for the sake of process. Organized without being bureaucratic.
HOW WE HIRE.

No HR screens. No culture-fit theater. No committees. Technical depth first — then alignment.

01
SEND A TRANSMISSION

Email disruptor@iron.codes with the role number in the subject. Tell us about the hardest system you've built and why it mattered. Not your life story — the one thing that proves you belong in this conversation.

02
TECHNICAL CONVERSATION

45 minutes. Architecture thinking, technical depth, system design instincts. Real problems — not hypotheticals. If you can't explain trade-offs, we can't build together.

03
REAL WORK CHALLENGE

Role-specific. For engineers: a live system design or a take-home build (48-hour window, respecting your time). For operations: a real scoping case. No trick questions — actual work.

04
FOUNDER CONVERSATION

45 minutes with the founding architect. Alignment, ambition, equity. We talk about where IRONCODE is going and whether you want to be the one who takes it there.

05
OFFER IN 48 HOURS

No committees. No consensus theater. Founder + one interviewer make the call. Offer within 48 hours of the final round. First day: pull request, not orientation slides.

"You'll join a founding team building a company that competes with Accenture — not by being bigger, but by being smarter. We use our own AI platform to diagnose organizations, we automate our own processes with agents, and we ship code, not slides. In 3 years, this is either the most respected AI architecture boutique in DACH, or a product company worth raising capital for. Either way, you helped build it from the first line of code."
IRONCODE // FOUNDING TEAM
READY? //

DROP A
TRANSMISSION.

Include the role number. We respond to serious operators only.

disruptor@iron.codes

// We don't do RFPs for talent.

// We don't do 6-round interview processes.

// We do conversations with people who know what they're worth

// and are done pretending otherwise.