AI Engineer
New
J
JoVEScientific research
Workable locations: IndiaFull-TimeMiddle
Salary not disclosed
Apply NowOpens the employer's application page
Job Details
- Experience
- 4+ years in software or ML engineering, including 2+ years shipping LLM, search or NLP features to production.
- Required Skills
- AWSDockerPythonElasticSearch
Requirements
- Have 4+ years of experience in software or ML engineering.
- Have 2+ years of experience shipping LLM, search, or NLP features to production.
- Demonstrate strong Python skills and software engineering practices, including testing, code review, and CI/CD.
- Have hands-on experience building RAG systems, including chunking, embeddings, hybrid retrieval, reranking, and prompt design.
- Have experience with a vector or search engine such as Elasticsearch/OpenSearch, Milvus, pgvector, Qdrant, or Pinecone.
- Have experience with LLM APIs such as Anthropic or OpenAI and open models such as Hugging Face or vLLM.
- Have built offline evaluations and used metrics to decide what ships.
- Have experience with agent frameworks, tool use, or building MCP servers.
- Be comfortable with AWS, Docker, and API design.
- Communicate clearly with non-engineers, including scientists and editors.
- Nice to have: experience with scholarly data, including PubMed, OpenAlex, Semantic Scholar, or PMC, or entity extraction for genes, organisms, and methods.
- Nice to have: multimodal work, fine-tuning embedding or reranker models, learning-to-rank, or large-scale indexing and vector quantisation.
Responsibilities
- Build hybrid retrieval using BM25, sparse and dense embeddings, and reranking over scientific papers.
- Improve matching of papers and protocol steps to JoVE videos and timestamps, accounting for organism, technique, and study type.
- Build RAG and agentic pipelines for cited synthesis, Methods Q&A, and troubleshooting, ensuring claims trace to sources.
- Extract method steps, reagents, and parameters from papers and link each step to the matching video chapter.
- Build evaluation sets with scientific editors and track relevance, citation accuracy, hallucination rate, and match quality.
- Ship and operate models and LLM services in production; monitor latency and cost per query and choose between hosted and open models.
- Help build APIs and an MCP server for AI assistants, ELNs, and publisher widgets to use JoVE methods data.
- Measure the search, video click, and watch funnel with product and use the results to improve ranking.
View Full Description & ApplyYou'll be redirected to the employer's site
