We are not hiring for this role right now. Send your CV anyway: we read every application and get in touch when a position opens.
The role
Build the AI features of id3, our software suite for drilling, and run them in production:
- the id3 AI Assistant, an LLM agent that answers drilling performance questions across wells, using tools over our drilling data and retrieval from documents;
- services that turn unstructured documents, such as daily drilling reports, into structured data;
- an agent that reviews data quality on live wells, combining deterministic checks with an LLM that must cite its sources, and a person who approves every fix.
You will work with our developers, drilling experts and clients, from the first prototype to a reliable service.
Our stack: Python with FastAPI and Pydantic; LangChain and LangGraph; Anthropic Claude, OpenAI and Azure OpenAI models; Qdrant for retrieval; PostgreSQL, Redis, Polars and DuckDB; everything in Docker on Kubernetes, on AWS and Azure.
What you will do
- Design and build LLM agents and the tools they use to query drilling data and KPIs.
- Build retrieval over documents and data: ingestion, chunking, embeddings and search.
- Extract structured data from unstructured documents.
- Write evaluations, and track the quality, cost and latency of AI features in production.
- Ship AI features as tested FastAPI services through CI/CD.
- Follow new models and techniques, and help the team decide what to adopt.
- Write technical documentation in English.
What we look for
- Strong Python.
- Hands-on experience building applications on top of LLMs: prompting, tool use, retrieval (RAG) and agents.
- A practical understanding of how LLMs behave: context limits, structured output, evaluation, cost and latency.
- Experience with web APIs (FastAPI or similar) and SQL.
- Confidence with Git, automated tests and working in a team.
- Good written and spoken English.
Bonus points for:
- LangGraph or LangChain, and vector databases such as Qdrant.
- The Anthropic, OpenAI or Azure OpenAI APIs.
- Machine learning on time-series or sensor data.
- Docker and Kubernetes, on AWS or Azure.
- Interest in the energy sector or in drilling data standards such as WITSML.
Why join kwantis
-
Projects around the world
Clients on every continent and projects in 30+ countries, in energy, construction and transportation.
-
A real career path
From Associate to Partner. Challenging international projects bring growing execution and managerial responsibility, and build your hard and soft skills.
-
A permanent contract
Full-time roles come with a permanent contract ("contratto a tempo indeterminato"), a competitive salary with performance bonuses, and monthly meal vouchers.
-
Flexible, hybrid work
Flexible hours, remote work from anywhere in Italy, and time for your life outside work. We meet in our office in the centre of Milan a couple of days a week.
-
An international team
Colleagues from more than 7 countries, and English as our everyday language.
-
A young, curious team
An agile, inclusive team that values curiosity and new ideas, and makes time for team building, go-kart races, climbing sessions or a beer.