Senior Software Engineer, - Information Retrieval
BenchSci
You Will:
- Design, implement, and optimize Conversational AI retrieval and evaluation systems in collaboration with data engineers, product managers, and scientists. This includes technical investigations, solution design, scalable implementation with automated tests, and participation in code reviews.
- Troubleshoot and resolve retrieval and embedding-related issues in a timely manner.
- Establish and uphold high standards forretrieval quality, evaluation rigor, and team culture, ensuring outcomes align with scientific and business goals.
- Be given an unmatched opportunity for accelerated growth and learn from a team working on cutting-edge IR + LLM applications in biomedicalresearch.
- Work on projects that directly support some of the largest pharmaceutical companies in the world, improving their ability to discover and apply biomedical knowledge.
- Tackle difficult IR/LLM challenges and contribute new ideas and perspectives to advance our retrieval platform.
You Have:
- Deep Expertise in Information Retrieval (IR): Strong expertise in information retrieval techniques (e.g., BM25, dense retrievers, hybrid methods) and an ability to optimize them in production.
- Evaluation Expertise: Hands-on experience with IR evaluation metrics (e.g., Precision, Recall, MAP, MRR).
- Hands-on Experience with LLMs and Embeddings: Familiarity with LLMs and vector embeddings, including fine-tuning or adapting models for domain-specific retrieval tasks.
- Proficiency in Python: High proficiency in Python, with experience building production-quality data or ML systems.
- System Design Skills: You can architect and build scalable information retrieval systems and understand the trade-offs of different design choices.
- 4+ years of professional software engineering experience, ideally in roles bridging ML/IR and backend development.
Nice to haves, but not mandatory qualifications:
- Infrastructure and Ecosystem: Exposure to cloud ecosystems (Google Cloud, BigQuery, Spanner, Vertex AI).
- Applied Machine Learning: You have experience tuning embedding models, training and evaluating reranker models, and are comfortable with A/B testing and online evaluation.
- Data and Knowledge Graphs: You have experience with graph databases and knowledge graph-based retrieval.