Internship position for World Models for Software Architecture (Constructor Fabric)

About the Project

Constructor Fabric turns a company's informal knowledge into production software through a pipeline of composable capability units called gears : requirements → architecture → product fit → framework (G1) → application → runtime. We are building a World Model over that pipeline — a model that does not just generate code, but predicts the consequences of an architectural decision: total cost of ownership, unintended side effects, latency and failure behaviour, and whether a proposed composition is even admissible.

The central object is not a digital twin of the application but its formal architectural skeleton : gear contracts (GearSpec), a typed attributed hypergraph of the application (AppGraph), a composition algebra that defines which assemblies are legal, and several semantic projections (types, protocols, resources, security) over which properties are proved, refuted by counterexample, or reported as an explainable gap . All of it is written in a domain-specific language whose syntax trees and graphs we keep in a projectional editor (JetBrains MPS) — so the DSL is not a convenience layer, it is the thing that defines the model's state space and the boundary of what the model is allowed to propose.

Internship Role

As a Research Intern (Machine Learning for Software Engineering), you will support the data and modelling work behind the World Model — mining code, building datasets, and running baseline experiments. We are looking for students who are keen to collaborate with industrial companies working with AI, open to out-of-the-box topics, and interested in a long-term collaboration.

Key Responsibilities

• Mine open-source repositories and extract architecture, dependency, and interface graphs from code.

• Help build and clean (specification → architecture) datasets, including synthetic project generation.

• Implement and evaluate baseline models (heuristic search, graph ML, or LLM generation) for single pipeline steps.

• Benchmark models against shared datasets and metrics to compare approaches.

• Build evaluation harnesses and track experiments reproducibly.

• Assist with documentation and, where relevant, publication activities.

Required Qualifications

• Currently enrolled in a BSc/MSc in Computer Science, Data Science, Mathematics, or similar.

• Experience with Python and ML frameworks (PyTorch, scikit-learn).

• Interest in machine learning for code, software engineering, or formal methods.

• Good English and teamwork skills.

Desired Qualifications

• Predictive models/time-series/etc.

• LLM tooling.

• Familiarity with graph machine learning (PyTorch Geometric or DGL).

• Interest in formal methods, program or static analysis, or working with code structure (e.g. AST manipulation, tree-sitter, or repository mining).

Application Details

Please submit:

• CV

• Short motivation letter (max. 1 page) explaining your interest and relevant skills

• Transcript of records (GPA obligatory)

Find Jobs in Germany on Arbeitnow