Description
AI Systems Architect - Multi-Agent & RAG Platforms
Role Overview
We are seeking an experienced AI Systems Architect to design and build an advanced, scalable AI platform leveraging customized Retrieval-Augmented Generation (RAG) and multi-agent (agentic swarm) architectures. The ideal candidate will lead the development of intelligent, modular AI agents capable of collaborative problem-solving across Governance, Risk, and Compliance (GRC) domains, with seamless integration into compliance platforms such as EasyCompliance.
Key Responsibilities
- Develop and optimize customized RAG pipelines tailored for enterprise GRC use cases.
- Build modular agent frameworks allowing dynamic addition of new skills, roles, and tools.
- Architect memory-driven agent systems, including shared memory, long-term context retention, and vectorless or hybrid memory approaches.
- Enable interoperability with multiple AI models (open-source and proprietary), ensuring flexibility in model selection and deployment.
- Deploy and manage AI models across local, cloud, or hybrid environments with performance and cost optimization.
- Integrate AI-generated outputs into compliance systems (e.g., EasyCompliance), including automated evidence generation and embedding.
- Design orchestration layers for agent coordination, task routing, and workflow execution.
- Ensure security, data governance, and auditability in AI workflows, especially for GRC-related outputs.
- Collaborate with stakeholders to translate compliance requirements into AI-driven solutions.
Required Skills & Qualifications
- Strong experience with LLMs, RAG architectures, and prompt engineering.
- Hands-on experience building or working with multi-agent systems (e.g., LangGraph, CrewAI, AutoGen, Semantic Kernel, or custom frameworks).
- Deep understanding of memory architectures in AI systems (vector databases, vectorless approaches, knowledge graphs, or hybrid memory systems).
- Experience deploying and managing open-source and commercial models (e.g., LLaMA, Mistral, GPT, Claude).
- Proficiency in Python and AI/ML frameworks (LangChain, Haystack, or similar).
- Experience with orchestration, distributed systems, and API design.
- Familiarity with embedding pipelines, retrieval systems, and document processing.
- Knowledge of containerization and deployment tools (Docker, Kubernetes).
Preferred Qualifications
- Experience with GRC platforms, compliance automation, or audit systems.
- Understanding of knowledge graphs and structured reasoning systems.
- Experience building tool-using agents (function calling, plugins, external integrations).
- Familiarity with vector databases (FAISS, Pinecone, Weaviate) and alternatives.
- Background in system architecture for enterprise-grade AI platforms.
- Exposure to security, risk, and compliance frameworks (ISO, SOC2, NIST).
Key Deliverables
- Fully functional multi-agent AI system with expandable roles and capabilities.
- Production-ready RAG pipeline integrated with GRC workflows.
- Memory-enabled agent ecosystem with shared and persistent context.
- AI-powered compliance evidence generation integrated into EasyCompliance.
- Scalable deployment architecture supporting multiple AI models.
Nice-to-Have Traits
- Strong problem-solving mindset with a systems-thinking approach.
- Ability to work in fast-evolving AI environments and evaluate new tools quickly.
- Experience in building extensible AI platforms rather than single-use solutions.
(ref:hirist.tech)
