Experienced AI Platform Engineer

We're Optivaize, a Dutch AI company that builds real AI-powered products — chatbots, intelligent platforms, and client-facing tools that leverage LLMs at their core. We don't do demos or prototypes that die in staging. We build production systems that clients use every day. We're looking for an AI Platform Engineer who understands how to build with LLMs, design robust AI architectures, and ship full-stack products.

WHAT YOU'LL DO
– Design and build AI-powered chatbots and agents using OpenAI, Claude, Gemini, and open-source models
– Build RAG (Retrieval Augmented Generation) pipelines: vector stores, embeddings, chunking strategies
– Architect and implement multi-agent systems, tool-calling flows, and LLM orchestration layers
– Engineer prompts systematically: you understand few-shot, chain-of-thought, and structured outputs
– Build full-stack platforms (React/Next.js frontend + Node.js or Python backend) that wrap AI capabilities
– Design scalable API layers between AI services and client-facing products
– Integrate AI into automation workflows (n8n and custom pipelines)
– Monitor AI system performance: latency, token costs, accuracy, and continuously optimize
– Stay current on model releases, fine-tuning options, and the evolving AI tooling landscape

STACK YOU'LL WORK WITH
OpenAI / Claude API | LangChain / LlamaIndex | Vector DBs | Next.js / React | Node.js / Python | PostgreSQL | n8n | Docker | Supabase / Pinecone

WHAT WE'RE LOOKING FOR
– Required 3+ years full-stack development with at least 1–2 years building with LLM APIs in production
– Required Hands-on experience with OpenAI or Anthropic APIs, not just wrappers, real production use
– Required Strong understanding of RAG, embeddings, and vector search (Pinecone, Qdrant, pgvector, etc.)
– Required Full-stack capability: React/Next.js + backend (Node.js or Python) + database design
– Required Ability to design system architecture, you can whiteboard an AI platform from scratch
– Preferred Understanding of AI cost management, you think about token efficiency, caching, model selection
– Preferred Experience with LangChain, LlamaIndex, or building custom orchestration frameworks
– Bonus Familiarity with fine-tuning or working with open-source models (Ollama, HuggingFace)

YOU'LL THRIVE HERE IF YOU...
– Get genuinely excited when a new model drops, you read the paper and test it immediately
– Understand that prompting is engineering, you approach it with rigor, not guesswork
– Can architect a full product, not just write isolated functions
– Think about cost and performance together, you don't just build, you build efficiently
– Communicate clearly, work async, and deliver without constant supervision

DETAILS
Salary range: $900 to $1600 per month
Type: Full-time contractor (with potential for permanent employment)
Hours: Flexible but with minimum 4 hours overlap with Central European TIme (UTC+1)
Location: Remote (Philippines)
13 Month: Paid
Probation: 30-day paid trial period