Bioinformatics Engineer - Zurich

Bioinformatics Engineer - Zurich

About Us

Revonode builds frontier AI intelligence for life-science business development. Deal teams at pharma corporate development groups and life science funds use our platform to run source-bound diligence on drug assets — across trial registries, patents, literature and regulatory records. In hours rather than weeks.

Our role is to help our partners see the full potential of a treatment, and to act on it. 

We are a dynamic and fast growing team of ETH-trained scientists and engineers, with prior operating experience across research, AI and biotech.

Your impact

You will own the high-fidelity biological data layer underneath our machine learning and AI intelligence. Moving beyond raw data ingestion, you will create curated datasets so that model training is consistently grounded in high-quality, standardized, version-controlled biological data.

By harmonizing disparate public datasets and internal data into a coherent representation, you will unlock the information that fuels our mission: getting the right treatments to the patients who need them, faster.

What you will do

• Build and own Revonode's biological dataset, applying bioinformatics best practices to the ingestion and harmonisation of complex data, so that model training is grounded in high-quality, version-controlled, and coherently integrated datasets.
• Curate and harmonise data from public sources (Open Targets, UniProt, Ensembl, ChEMBL, Reactome and others), implementing rigorous strategies for identifier mapping, version drift, and conflicts between databases.
• Use LLMs to extract structured information from papers, patents, and internal documents at scale, verifying the outputs with a biologist's eye.
• Run exploratory and hypothesis-driven analysis across genomics, proteomics, and target-level data to inform what the system should learn.
• Feed curation insights back into the design of the agentic flow and ML models.
• Document curation rules and data provenance so that decisions are reproducible and the wider team can build on them.

Skills and qualifications

Essential

• A background in biology (a degree in biology, bioinformatics, computational biology or a related field, or equivalent hands-on experience), with enough domain knowledge to recognize when data is wrong, not just malformed.
• Demonstrable experience in bioinformatics data analysis across multiple analytic methods such as, genomic, proteomic, or target-level datasets.
• Working familiarity with the major public biological databases.
• Hands-on experience using LLMs for information extraction, summarization, or classification.
• Proficiency in Python sufficient to manipulate data, script analyses, and work with APIs.

Nice to have

• Experience with graph databases (FalkorDB, Neo4j) or knowledge-graph modelling of biological entities.
• Exposure to ML evaluation concepts such as label quality, inter-annotator agreement, and train/test separation.
• Previous experience in a drug discovery or biotech environment.
• Familiarity with cloud data infrastructure.

Culture and values

You will own an area outright and see it reach production quickly. You will have daily interactions with the leadership team, as priorities shift as we scale. Some of the needed tooling you will build yourself, because it does not exist yet. We think this is the most exciting place to be. 

We are guided by three values: 

Rigour : Getting the science right matters most to our customers. We are curious, creative and deliberate in how we approach a problem, uncompromising about the quality of the answer. 

Transparency : We communicate openly, share what we know, say when we are unsure, and help each other without being asked. A team only works when everyone understands the whole picture. 

Ownership : Everyone at Revonode owns an outcome, not a task list. We take the initiative, follow through, and hold ourselves accountable for the impact of our work. 

Details

Full-time, on-site in Zurich.

Interested? Apply here, or message us directly with one paragraph on a dataset you have curated and what you found wrong in it.