AI/Software Internship - #54699

What To Expect
Start date: September 2026 

Location: Tilburg

Duration: 6 months

Time type: Full-time (40 hours a week)

NOTE: YOU MUST BE ENROLLED IN EDUCATION DURING THE ENTIRE INTERNSHIP TO BE CONSIDERED.

The AI / Software Intern is responsible for helping move the GOBI team's data and application stack — which today runs the dashboards, web applications, and data pipelines behind Tesla's remanufacturing operations in Lathrop, Tilburg, and Buffalo — toward a modern agentic data platform. Through this role we will get some help maintaining our data pipelines and increasing their reliability, as well as integrating both existing and new dashboards and visualizations with AI agents to extend them and generate automatic insights and reports.

The intern will build a semantic / metrics catalog over the existing cache schemas, develop an agent loop on inference endpoints, wrap operational APIs as MCP tools so agents can both read and act, integrate the result into the existing frontend, and put evals and observability in place before any agent reaches production. They will work on the GOBI ETL, visualizations, backend and frontend, and collaborate with other existing stakeholders and teams.

What You'll Do

Build a unified semantic / metrics catalog over the existing cache schemas by walking current dashboards, ETLs and ELTs

Evolve the team's text-to-SQL design into an agent loop with tool calling, multi-step planning, self-correcting SQL and citations back to source tables. Wrap operational APIs as MCP tools so agents can both read and act

Integrate the resulting agent and data products into the existing React + Vite + Tesla Design System frontend

Build a fixed eval set and unit tests, and wire up traces / accuracy / latency dashboards before anything is shipped to production

Contribute to the underlying FastAPI / SQLAlchemy services, Airflow DAGs and CI/CD pipelines as needed to support the above

Document the semantic layer, agent tools and evals so the rest of the GOBI team can extend them

What You'll Bring

Currently enrolled in Computer Science, AI Engineering, ML Engineering, Data Engineering or a similar field

Strong Python and SQL — able to read a non-trivial FastAPI service or a long ETL script and refactor across them

Hands-on experience with at least one modern data warehouse / query engine (Snowflake, Databricks, BigQuery, Trino/Presto, Vertica)

Has built something with LLMs beyond a chat wrapper — RAG, function/tool calling, an agent loop, or an MCP server

Working knowledge of CI/CD, Git, Docker and Kubernetes

Good written and spoken English, with the ability to communicate technical trade-offs clearly

Nice to have: dbt / Cube / Malloy, LangGraph / OpenAI Agents SDK / MCP, React + Vite, LLM evals & observability (Langfuse, Phoenix, Braintrust), or a time-series data analysis background

, Tesla