Responsibilities
End-to-end ownership of architecture, development, and go-live of AI agents and LLM applications
Architecture of agentic systems: multi-step workflows, tool use, retrieval-augmented generation, and model selection and switching based on data protection, cost, and accuracy
Establish evaluation and guardrails as standard: solid quality evidence before every go-live and traceability for auditors and regulators
Technical sparring for clients and Strategy: make-or-buy, effort, risk, compliance, and implementation options
Lead workshops with business teams and management and make complex systems understandable for every audience
Set engineering standards for agentic work: code reviews, workflows with AI coding agents, mentoring, and building knowledge in the team
Own security and data protection concepts for AI in regulated environments, including operation in the client's environment
Qualifications
5+ years of experience in software development, of which >2 years in a leading or architectural role
Deep hands-on practice with LLM applications in production: RAG, tools/agents, evaluations, and guardrails
Excellent TypeScript skills (e.g. React, NestJS) and experience designing APIs and system integrations
You work with AI coding agents (e.g. Claude Code, Codex, Cursor) as a matter of course and know how a team ships quickly and cleanly with them
You get to grips with unfamiliar system landscapes quickly and spot risks before they become problems
Confident in cloud and on-premise environments, integration into existing enterprise IT, and operating security-relevant systems
A strong ability to translate technical complexity for business stakeholders up to executive management
Willingness to work on-site with our clients
German at C1/C2 level, plus excellent English
Nice to Have
Experience in regulated industries such as banking, insurance, or the public sector with heavy compliance and audit requirements
Experience running open-weight models, fine-tuning, or building evaluation environments for agents
Open-source contributions, publications, or talks on agents and applied AI
What We Offer
Scope to shape things: direct influence on how we build, test, and run AI agents – and on portfolio and client projects
Work at the forefront of generative AI: the newest models and techniques, deployed where reliability is mandatory
Leadership responsibility for project teams and active mentoring of junior colleagues
The opportunity to build thought leadership (publications, conferences, talks)
Close collaboration with management and C-level clients
Long-term development prospects toward a partner role / tech lead / practice lead
Attractive, performance-based compensation and benefits
Case-team events & employee benefits (company fitness, mobility)
Don't tick every box? Results from your projects convince us more than a complete checklist. If you want to build AI solutions that actually run, we want to hear from you.
Send your CV to info@blackfield.ai. We look forward to your application!