SDE III Gen AI

About The Role

You will design, implement, and deploy production-ready generative AI applications serving millions of users. Key responsibilities include building advanced RAG pipelines with vector databases and hybrid search for sub-second responses; developing multimodal systems integrating text, vision, and audio; architecting scalable microservices on GCP; and optimizing LLMs via fine-tuning.

Gen AI Specific Responsibilities

• Fine-tune state-of-the-art language models for specific business use cases to improve accuracy and relevance.
• Design multi-agent AI systems using frameworks like LangGraph and Google ADK to orchestrate complex workflows.
• Implement advanced prompt engineering (Tree of Thoughts, ReAct) and build production-grade embedding systems handling billions of vectors.
• Create secure applications with robust safeguards against injection attacks while maintaining governance compliance.
• Optimize token usage for 50-70% cost reduction without compromising quality.
• Build real-time inference systems capable of sub-100ms latency and develop multimodal search across text, images, and documents.

We're Looking For

Candidates should have 5+ years in ML/AI with at least 2+ focused on Gen AI/LLMs. Expert-level Python skills (async, multiprocessing) are required alongside deep knowledge of PyTorch, Transformers, LangChain, and GCP/Kubernetes.

• Proven track record deploying search/discovery solutions using dense retrieval, bi-encoders, cross-encoder reranking, and hybrid BM25 + vector search for large SKU catalogs.
• Experience building multi-agent systems with tool-augmented agents for real-world workflows.

Nice to Have

Published research, multimodal AI experience, domain expertise in e-commerce/finance/legal, knowledge of AI safety principles, and familiarity with emerging technologies like MoE or NAS.