Remote, Remote, North America, Europe TGScope is looking for a research scientist for a project-based contract of about two months, fully remote, paid by the hour. The project: build detection of coordinated channel networks across 3 million public Telegram channels, using forwards, shared links, mentions and near-duplicate posts, and validate it against known networks. How Python is used: the whole project is Python. Data preparation, graph construction (NetworkX or igraph), near-duplicate detection and validation are written in Python on top of PostgreSQL, DuckDB and Parquet. Stack: Python, PostgreSQL, DuckDB, Parquet, Qdrant with multilingual-e5 embeddings, NetworkX or igraph. Background in network science, graph ML or computational social science Practical experience with community detection and near-duplicate text detection at scale Strong Python and SQL on tens of millions of rows Solid statistics Written English
You apply on Python.org Jobs, where this job was posted. Peter Nowak isn’t the employer and doesn’t check listings.