Backend Python Engineer; OSI-SIS-TS--GRAP

Job Description

You will join the Tools and Services section within the Scientific Information Service, the team responsible for the platforms that power scientific discovery and open access across the High Energy Physics community. Our systems include INSPIRE, the global literature hub for particle physics; SCOAP3, which funds and tracks open access publishing; CERN Analysis Preservation, which safeguards the analyses behind physics results;

and a tool facilitating access to CERN's institutional memory.

As a Backend Python Engineer you will design, build and operate the services that ingest, curate, search and expose millions of scientific records to a worldwide audience. You will work on high‑throughput data pipelines, REST APIs, search infrastructure and the integration of AI‑driven retrieval into production systems. Your code will be open source and used daily by researchers across hundreds of institutes.

As part of the Tools and Services team, you will drive the modernisation and evolution of the Scientific Information Service content platform. You will advance the backend services behind INSPIRE, SCOAP3 and CERN Analysis Preservation, expand Model Context Protocol services that open the team’s scientific content to AI clients, and evaluate the harmonisation of the INSPIRE backend with the newer back‑office stack.

• The core responsibility of this position is to drive the modernisation of the INSPIRE backend, expand the team’s Model Context Protocol services, and align the digitisation and archival services with the team’s technology stack, delivering a large‑scale migration to the Access to Memory (AtoM) service.

• Review the remaining custom harvests, approximately 1,000, and migrate those that should be retained to Apache Airflow, retiring or redesigning the rest.
• Work with curators to decide what to keep, and automate the pipeline where evaluation shows it adds value, keeping a human in the loop for decisions that require expert judgement.
• Evolve the Apache Airflow ingestion and curation pipelines that process scientific metadata and full text at scale, improving data quality, deduplication, enrichment and source coverage.

• Expand the team’s Model Context Protocol services across INSPIRE, the library and other team services, moving them from pilot to a consistent, documented and maintainable interface that opens curated scientific content to external AI clients.
• Develop and optimise the search and discovery functionality on Open Search that underpins these features, including relevance tuning and indexing strategies.

• Align the digitisation and archival services with the team’s technology stack, modernising them onto common frameworks and deployment tooling.
• Deliver and oversee the migration of 35,000 new archival items to the Access to Memory (AtoM) service across 2027, maintaining consistent metadata and data quality throughout.
• Develop and automate the connection between the connected services, enabling content to flow reliably across the digitisation and archival systems.
• Prepare the ingestion pipelines for new file ingestion, harmonising them with the team’s other applications.

• Design, document and deliver REST APIs consumed by internal applications, partner institutes and third‑party tools.
• Apply automated testing, code review, monitoring and CI/CD, and support production through containerised deployments and performance tuning.
• Contribute the team’s outputs back to the open‑source community to advance Open Science.

• Demonstrated experience building and operating backend services in Python in a production environment. Solid track record with web frameworks such as Django, Flask or FastAPI and with relational and document databases.
• Hands‑on experience designing REST APIs.
• Practical experience with search engines such as Open Search, including indexing and relevance tuning.
• Familiarity with containerised deployments, version control with Git and CI/CD workflows.
• Experience contributing to open‑source projects and working in a collaborative engineering…