Principal Solution Architect - AI Platform - Freelance

Context

The role is within the AI Platform team of a major corporate and investment banking division. This team serves as the core engineering engine driving the next generation of Artificial Intelligence in the banking sector. The team is currently leading the establishment of an Agentic Project Center of Excellence (CoE) focused on building an advanced ecosystem of autonomous, reasoning agents designed to navigate complex financial workflows. The position involves architecting a high-performance, scalable, and resilient infrastructure that enables Large Language Models (LLMs) and multi-agent systems to interact with enterprise data and tools.

Missions

• Define the long-term technical roadmap for the AI Platform, focusing on integrating Multi-Agent Systems (MAS) into the bank’s core infrastructure.
• Architect complex orchestration layers using LangGraph, Google ADK, and CrewAI to ensure seamless interaction between autonomous agents, toolsets, and enterprise data.
• Lead the adoption of the Model Context Protocol (FastMCP) to standardize agent interactions with internal tools and external data sources.
• Design architectural guardrails for agentic reasoning, including human-in-the-loop patterns, compliance checkpoints, and deterministic safety layers to manage LLM unpredictability.
• Oversee the design of highly resilient, event-driven architectures and containerized environments (Kubernetes/GCP) to support large-scale AI workloads.
• Act as the primary technical liaison between Data Science, Backend Engineering, and Business Stakeholders.
• Establish enterprise-wide standards for AI software craftsmanship, MLOps, and LLM evaluation frameworks.
• Mentor senior engineering leads.
• Translate business processes into technical implementations with a focus on minimal coupling and reusability.
• Enforce reuse of existing components by collaborating closely with functional and technical architects.

Tools & Environment

• Multi-Agent patterns, Microservices, Event-Driven Architecture (EDA), Distributed Systems, High-Availability Design
• Multi-Agent Orchestration (MAS)
• LangGraph, CrewAI, FastMCP
• Google AI SDK (ADK)
• Kubernetes, Google Cloud Platform (GCP)
• Large-scale distributed systems, MCP, SQL/NoSQL, API Management
• Vector Databases & Retrieval-Augmented Generation (RAG) Architectures
• Advanced CI/CD, Automated Testing (Cucumber/BDD), Observability
• AI Safety & Guardrail Implementation
• Compliance frameworks including GDPR, MiFID, and AI Act

Working Conditions

• Role type: Full-Time / Executive Level
• Experience required: 15+ years
• Location: Within a major corporate and investment banking environment
• High-visibility leadership position with responsibility for technical vision and architectural blueprinting of AI and Agentic systems in production environments