What it's about:

Our AI octo-workers automate complex knowledge work with verified quality. For this, they need clean, consistent, and reproducible data. That's precisely what you'll be responsible for: the end-to-end integration of the sources and services that supply our platform with reliable information. You'll transform unstructured customer data into something our LLM agents can work with. What you'll have built in 12 months

• Productive data and extraction pipelines, from unstructured sources to AI-ready services, that run reproducibly even when sources change.
• Proprietary tools and integrations against customer APIs, ticketing systems and internal systems, often as same-day or next-day prototypes.
• A knowledge base structure that measurably improves retrieval and embedding quality.
• Reproducibility and service standards upon which the rest of the team builds.

What you will do

• Translating customer requirements into functioning technical solutions together with our AI Solutions Managers.
• Extract, clean, and transform data for ingestion into our RAG pipeline.
• Build special functions that the core platform does not yet cover.
• Quickly iterate on customer feedback and then harden the solution into a stable service.

Who this is for:

You think end-to-end, from requirements to a stable service. You write production-ready Python and make technical decisions with LLM behavior in mind: embeddings, ingestion, chunking, and retrieval. You have experience building with RAG and LLM agents, in production or in serious side projects. You are AI-native: Claude, Cursor, or similar tools are part of your daily routine, and you can describe how they have changed your work. You can handle fuzzy, underdetermined problems where data and requirements are initially unclear. Must-haves

• You have delivered production-ready Python services.
• Docker.
• Hands-on experience with RAG and with LLM agents or agentic workflows, professionally or in substantial side projects.
• Text manipulation and data wrangling.
• Experience with Agentic Coding.
• Evidence of what you have built: public repos, a take-home or walkthrough of a private project.

Nice to have

• Data engineering background.
• CI/CD and Cloud (AWS, Azure).
• 
  
•