Vanguard is looking to solve the problem of building and deploying robust and scalable GenAI applications and RAG systems that can provide high-quality retrieval and ranking capabilities, while ensuring responsible AI practices and security governance.
Requirements
- Hands‑on experience building GenAI applications and RAG systems end‑to‑end.
- Strong proficiency in Python for ML/LLM development.
- Experience with vector databases (e.g., pgvector, Pinecone, Weaviate, FAISS) and embedding models.
- Knowledge of LLM frameworks (LangChain, LlamaIndex, Transformers, etc.).
- Strong understanding of cloud environments (AWS/Azure) and containerized deployments.
- Solid software engineering foundations — APIs, microservices, version control, testing, CI/CD.
- Experience with data pipelines, ETL/ELT, and processing unstructured data.
Responsibilities
- Architect, build, and deploy RAG pipelines, including chunking, embeddings, vector stores, retrieval, ranking, grounding, and evaluation.
- Design and implement Graph RAG solutions leveraging knowledge graphs for multi‑hop reasoning and structured retrieval.
- Build robust, scalable ML/LLM services using Python (and Java where applicable) with well‑designed APIs and microservices.
- Develop data processing pipelines for ingestion, transformation, metadata extraction, and indexing.
- Implement observability, monitoring, evaluation harnesses, automated testing, and CI/CD for GenAI services.
- Optimize retrieval quality, response accuracy, latency, and cost across model + retrieval layers.
- Apply responsible AI, security, and governance practices for LLM systems (e.g., content filtering, guardrails, model monitoring).
Other
- 3+ years of experience as an ML Engineer, AI Engineer, or similar role.
- Excellent communication skills and ability to work in cross‑functional teams.
- Ability to produce clear documentation, design specs, and operational runbooks for all delivered components.
- Vanguard is not offering visa sponsorship for this position.
- A builder who loves engineering elegant, reliable systems that scale — someone who understands both machine learning and strong software engineering practices, is passionate about GenAI, and is excited to push the boundaries of RAG and Graph RAG capabilities.
