Solution Architect - LangGraph & Agentic AI

Belmont Lavan Ltd; Stuttgart; Vor Ort; EUR 120.000 - 170.000; Vollzeit

Zusammenfassung

Belmont Lavan Ltd in Stuttgart seeks an experienced Solution Architect to lead the architecture of LangGraph-based AI platforms and applications, ensuring secure, production-ready designs that scale enterprise-wide.

You will collaborate with business and technology stakeholders to identify high-value AI opportunities, translate them into robust architectures, govern deployments, and mentor engineering teams.

Qualifikationen

• Significant experience in solution architecture or AI architecture.
• Hands-on with LangGraph-based AI applications or agentic workflows.
• Strong understanding of LLM application architectures and enterprise security.

Aufgaben

• Lead design of enterprise AI agent and agentic workflow solutions.
• Translate business requirements into secure, scalable architectures.
• Define reusable architecture patterns and govern AI deployments.

Kenntnisse

LangGraph

LLMs

Agent architectures

Cloud architectures

Security & governance

Observability

Python

Stakeholder leadership

Tools

LangChain

Kubernetes

APIs

Jobbeschreibung

We are looking for an experienced Solution Architect with hands‑on experience designing and deploying LangGraph‑based AI solutions to lead the architecture of enterprise agentic AI platforms and applications.

You will work with business and technology stakeholders to identify high-value AI opportunities and translate them into secure, scalable, and production‑ready architectures.

The role combines AI architecture, enterprise integration, cloud engineering, agentic AI, security, governance, and stakeholder leadership.

You will be expected to understand LangGraph at a practical level and be able to challenge architectural decisions, guide engineering teams, and ensure that AI solutions can operate reliably at enterprise scale.

Requirements

AI Solution Architecture

• Lead the architecture and design of enterprise AI agent and agentic workflow solutions.
• Design LangGraph‑based architectures for single‑agent and multi‑agent applications.
• Translate business requirements, processes, SLAs, security requirements, and technical constraints into solution architectures.
• Evaluate architectural alternatives and document key technical decisions and trade‑offs.
• Define reusable architecture patterns for agentic AI solutions.

Enterprise Agent Architecture

• Design architectures incorporating:
• LLMs
• LangGraph
• RAG
• Enterprise data
• APIs and business systems
• Workflow engines
• Human approval processes
• Observability
• Security and governance
• Define appropriate boundaries between AI reasoning and deterministic business logic.
• Design state management, persistence, recovery, and long‑running agent workflows.
• Determine when to use single‑agent, multi‑agent, or conventional application architectures.

Cloud and Platform Architecture

• Design scalable AI application architectures on AWS, Azure, or GCP.
• Define compute, networking, storage, API, security, and platform requirements.
• Design architectures suitable for enterprise‑scale production workloads.
• Evaluate cloud services and AI platform capabilities based on performance, security, scalability, and cost.
• Work with platform engineering and DevOps teams to establish deployment standards.

Integration Architecture

• Design integration between AI agents and enterprise applications, APIs, databases, and SaaS platforms.
• Define secure mechanisms for agent tool access and business‑system interactions.
• Design authentication, authorisation, secrets management, and access‑control approaches.
• Ensure AI‑driven actions are traceable, auditable, and appropriately governed.

AI Security and Governance

• Establish security and governance principles for enterprise AI agents.
• Address risks including:
• Prompt injection
• Data leakage
• Unauthorised tool usage
• Excessive agent permissions
• Inaccurate or unsafe actions
• Sensitive‑data exposure
• Define appropriate human‑in‑the‑loop controls.
• Ensure solutions comply with organisational security, privacy, regulatory, and responsible‑AI requirements.

AI Evaluation and Observability

• Define architecture for AI application monitoring and observability.
• Establish approaches for evaluating agent accuracy, reliability, latency, cost, and task completion.
• Define appropriate logging, tracing, metrics, and alerting.
• Establish operational processes for monitoring and continuously improving production agents.

Stakeholder and Technical Leadership

• Work directly with senior business and technology stakeholders to define AI strategies and roadmaps.
• Lead architecture workshops and technical design sessions.
• Communicate complex AI concepts and architectural trade‑offs to technical and non‑technical audiences.
• Provide technical direction to AI engineers, developers, data teams, and platform engineers.
• Review solution designs and ensure alignment with enterprise architecture standards.
• Mentor engineering teams and promote reusable AI architecture patterns.

Required Experience

• Significant experience in solution architecture, software architecture, AI architecture, or a related role.
• Hands‑on experience designing and deploying LangGraph‑based AI applications or agentic workflows.
• Strong understanding of LLM application architectures.
• Experience with enterprise AI/ML solutions in production.
• Strong understanding of RAG, tool calling, agent orchestration, and human‑in‑the‑loop patterns.
• Strong experience with at least one major cloud platform: AWS, Azure, or GCP.
• Strong understanding of enterprise integration patterns and APIs.
• Experience with security, governance, observability, and operational requirements for production systems.
• Strong technical understanding of Python and modern software engineering practices.

Desirable Experience

• LangChain / LangSmith
• Multi‑agent architectures
• Enterprise RAG platforms
• Vector databases
• Kubernetes
• Event‑driven architectures
• Microservices
• Infrastructure as Code
• CI/CD
• MLOps / LLMOps
• AI security
• Responsible AI
• Large‑scale enterprise transformation
• Experience working directly with senior client stakeholders