The Role

We're looking for an AI Engineer who wants to work on production AI systems that actually ship to users. You'll evolve our AI/ML infrastructure - from document ingestion pipelines to RAG retrieval to knowledge graph extraction.

This is a systems-focused AI role. You'll work across the full stack: designing embedding strategies, optimizing vector search, building agent workflows, and ensuring our AI features are fast, accurate, and reliable in production.

What You'll Work On

  • Intelligent Document Understanding
  • AI-Powered Workflows
  • Knowledge Graph & Structured Data
  • Workspace Intelligence
  • Production AI Infrastructure

Our Stack

Python React Langchain Pgvector ArangoDB FastAPI Azure

What We're Looking For

Required

AI/ML Fundamentals

  • Strong understanding of embedding models, vector search, and similarity metrics
  • Experience building RAG systems (not just OpenAI API wrappers)
  • Familiarity with evaluation metrics (MRR, NDCG, precision@k, F1)
  • Understanding of prompt engineering and LLM limitations

Python Engineering

  • Production Python experience (async/await, type hints, pydantic)
  • Comfortable with numpy, pandas, or similar data manipulation libraries
  • Experience with testing and CI/CD
  • Ability to profile and optimize performance bottlenecks

Systems Thinking

  • Understanding of latency/cost tradeoffs in production AI systems
  • Ability to debug complex async pipelines
  • Comfort reading research papers and implementing new techniques

Pragmatism

  • You ship code that works, not just research experiments
  • You balance quality with velocity
  • You're comfortable with "good enough" when appropriate
  • You can explain technical tradeoffs to non-technical stakeholders

Nice to Have

  • Experience with pgvector, Pinecone, Weaviate, or similar vector databases
  • Knowledge graph experience (entity extraction, relation extraction, ArangoDB/Neo4j)
  • Workflow automation or business process automation background
  • Experience building multi-agent systems or tool-using LLMs
  • Familiarity with reranking models (Cohere, BGE reranker)

What Makes This Role Different

You'll Work on Real Problems

  • Not just toy datasets - real business workflows with messy documents, complex forms, and diverse data types
  • Production systems where latency, accuracy, and cost all matter
  • Multi-tenant architecture serving diverse industries and use cases

You'll Build Product, Not Just Features

  • AI isn't a feature here - it's foundational to how workspaces adapt and workflows execute
  • You'll design how AI enables customization (smart templates, workflow suggestions, auto-configuration)
  • Your work directly impacts whether users can replace 3-4 different tools with one intelligent platform

Our workplace rocks

  • We are a product team of 8 devs, 2 UX/UI designers.
  • We are established, we aren't struggling month to month.
  • Great office space with; lunch, a fridge that magically keeps filling itself, and top tier work equipment.

You'll Solve Diverse AI Problems

  • Not just RAG: document generation, form extraction, workflow automation, template matching, entity resolution
  • Work across graph databases, vector search, LLM agents, structured data, and traditional ML
  • Exposure to full-stack development (FastAPI backend, React frontend, ArangoDB graph queries)
  • Balance research (reading papers, trying new models) with engineering (making it production-ready)

You'll Have Autonomy

  • We hire smart people and trust them to make good decisions
  • You'll propose and drive AI initiatives based on user needs and product vision
  • Flexible about tools and approaches - use what works for the problem at hand

Our Engineering Culture

Code Quality Matters

  • We have a 5-layer backend architecture with clear separation of concerns
  • Strong typing (Pydantic models, TypeScript)
  • Comprehensive test coverage with pytest and automatic cleanup
  • Pre-commit hooks and CI/CD to catch issues early

Documentation is First-Class

  • We maintain extensive docs (architecture, conventions, examples)
  • Code reviews focus on clarity and maintainability
  • We write context documents for complex systems

We Delete Code

  • "No code is the best code" - we're not precious about our work
  • We refactor aggressively when patterns emerge
  • We delete unused features rather than letting them rot

We Ship

  • Blue-green deployments with zero downtime
  • Comprehensive monitoring and error tracking
  • Fast iteration cycles - changes go live in days, not months

We're looking forward to hearing from you!

Etain is ISO/IEC 27001–certified, demonstrating our commitment to rigorous, audited information-security practices. We are also GDPR compliant and are actively working toward achieving SOC certification.

Certified
ISO 27001 ISO 27001
Compliant
GDPR GDPR