AI Software Engineering Intern at Cohesity · Santa…

Summary: Cohesity is the leader in AI-powered data security, serving over 13,600 enterprise customers. The AI/ML intern will work on improving retrieval and generation systems to enhance answer quality and user satisfaction in enterprise applications. Responsibilities: - Design and benchmark hybrid retrieval pipelines combining dense semantic search with sparse keyword techniques (BM25, TF-IDF). Experiment with fusion strategies — RRF, weighted merging — and evaluate their impact on answer quality across diverse query types - Investigate chunking strategies, context window construction, and reranking approaches to maximize answer grounding. Build automated pipelines to detect and reduce hallucinations, with exact source attribution as a first-class requirement - Own end-to-end quality for a QA system serving enterprise users — from query understanding and document retrieval through answer generation and post-processing. Identify failure modes and drive measurable improvements across precision, recall, and user satisfaction proxies - Build eval harnesses using frameworks like RAGAS or custom metrics (context precision, answer relevance, faithfulness) to continuously measure pipeline quality. Use eval results to drive data-informed decisions on model selection, prompt design, and retrieval configuration Required Qualifications: - Pursuing a graduate degree in ML, CS, or related field - Hands-on experience building RAG or QA pipelines - Strong Python skills; comfortable working with data at scale - Familiarity with vector databases and embedding models - Understanding of information retrieval fundamentals Required Skills: Machine Learning, Python, Information Retrieval, Vector Databases, Embedding Models, Retrieval-Augmented Generation (RAG), Question Answering Pipelines, Semantic Search, Keyword Search Techniques, Evaluation Metrics for NLP, Prompt Design