Why we're hiring:
Delvo is building the agentic intelligence layer for strategic procurement. We need an AI Engineer to design reliable LLM workflows and agent systems that work in real enterprise environments, not just demos.
Responsibilities:
- Build LLM workflows for retrieval, tool use, structured outputs, and multi-step reasoning
- Design agent orchestration with tools, control flows, retries, safety checks, and graceful degradation
- Develop evaluation & monitoring infrastructure with golden datasets, online metrics, tracing, and regression detection
- Implement enterprise integrations with ERPs and data sources with strong observability
- Optimize cost & latency via caching, streaming, batching, and intelligent model routing
- Collaborate with design and forward-deployed engineers to ship AI that works for real users
Ideal candidate:
- Proficient in TypeScript & Python across backend and product-adjacent work (Next.js, workers, APIs)
- Hands-on experience with LLMs: RAG, function calling, tools, structured parsing, guardrails, streaming
- Familiar with modern AI stack: Vercel AI SDK, OpenAI/Azure, embeddings, vector stores, Langfuse or similar
- Strong evals & quality mindset: define tasks, gold data, and success criteria; prevent regressions systematically
- Bonus: data pipelines, ERP integrations, procurement domain knowledge
What you'll get:
- Competitive salary + meaningful equity — real ownership in a company you'll shape
- Top-Spec MacBook Pro
- AI-first tooling — unlimited tokens, best-in-class tools, investment in your setup
- Deep technical growth in agentic systems, reliability engineering, and production AI
- Direct founder access — small team, visible outcomes, no layers
- Delta Campus Berlin, Neukölln — office in the heart of Berlin's startup ecosystem
- Path to AI/ML leadership as we scale from founding team to engineering org