Description

You will design and build a Model-as-a-Service platform enabling non-experts to construct AI solutions via drag-and-drop components.

Responsibilities

• Architect and optimize end-to-end Retrieval-Augmented Generation (RAG) pipelines, including advanced chunking strategies and vector database management.
• Develop and maintain MCP (Model Control Protocol) libraries, clients, and servers to integrate diverse data sources with the AI engine.
• Build and curate a repository for Agentic AI, allowing users to select or construct custom agents for specialized tasks.
• Manage and optimize one of the largest on-premise GPU farms in the U.S. banking sector, comprising 500+ Nvidia nodes.
• Integrate AI deployment pipelines with enterprise CI/CD tools like Jenkins and Ansible while implementing corporate guardrails within Model Risk Management (MRM) frameworks.

Required Skills

• Expert-level Python with deep, hands-on production experience.
• 10-15+ years of experience building production-grade platforms for developers or business units.
• Extensive data engineering background in massive data ingestion and processing.
• Deep understanding of vector databases, inferencing, and advanced RAG chunking strategies.
• Experience mimicking cloud capabilities (AWS/Azure) within strictly on-premise environments.
• Proficiency with DevOps tools, specifically Jenkins and Ansible for automated deployment.
• Proven ability to translate business use cases into scalable platform services.

Preferred Skills

• Experience working with large-scale GPU farms and high-volume data environments.
• Current knowledge of agentic frameworks and rag-less inferencing techniques.

Key Skills

Education

Any Gradute