Architect AI Engineer

Exadel is an AI-first global tech company with 25+ years of engineering experience, employing over 2,000 professionals and managing 500+ active projects for Fortune 500 clients such as HBO, Microsoft, Google, and Starbucks. The client is a global cloud software leader in corporate compliance and investor relations, supporting thousands of companies with transparency, simplified reporting, and stakeholder communication.

• 10+ years of overall experience, with at least 4 years in a dedicated Architect, Lead, or Principal role designing distributed enterprise SaaS platforms
• 5+ years of experience architecting on major cloud providers, with direct deployment experience in Azure AI Foundry, AWS Bedrock, or GCP Vertex AI
• Expert-level engineering foundation in Python (3.13+ preferred), asynchronous programming, and microservices architecture
• Proven experience with containerized deployment, MLOps pipelines, and operational realities of production AI (failure handling, fallback patterns)
• 3+ years of hands-on experience building and deploying production-grade LLM applications
• Minimum of 1.5 years specifically building autonomous agents and stateful AI interactions
• Deep hands-on experience with LangGraph (preferred), LangChain, LlamaIndex, Semantic Kernel, or similar graph-based orchestration frameworks
• 2+ years of experience with Vector Databases (Pinecone, Weaviate, Qdrant) or Graph Databases (Neo4j)

Nice to have:

• Active experimentation or professional implementation of Model Context Protocol (MCP) or A2A messaging standards
• Proven track record of designing knowledge-graph-grounded AI applications
• Active contributions to the LangChain/LangGraph ecosystem or other major AI frameworks
• Direct experience implementing Prompt Shields, Guardrails for Bedrock, or AI Content Safety gateways

• Architect multi-agent collaboration patterns (Supervisor-Worker, Planner-Executor, Swarm, Human-in-the-loop) capable of planning, reasoning, and safe execution
• Build the core platform supporting model orchestration across multiple providers (OpenAI, Anthropic, Bedrock, Azure, Vertex), including model gateways, prompt versioning, and execution guards
• Define and implement protocols for agent-to-agent (A2A) and agent-to-system communication, establishing Model Context Protocol (MCP) standards for tool-use
• Collaborate with frontend teams to implement AG-UI or ChatKit protocols, creating "UI tools" that allow AI to interact directly with the application interface
• Implement and optimize retrieval architectures, transitioning from basic semantic search to hybrid, vector, and knowledge-graph-enhanced retrieval (GraphRAG)
• Design safe "leave-and-layer" modernization patterns, exposing legacy enterprise functionality to AI agents through wrapper APIs, event-driven workflows, and integration gateways
• Architect long-term memory solutions (e.g., Graphiti, Cognee) to persist facts, observations, and user preferences across sessions
• Implement governance including prompt/output validation, policy-as-code enforcement, role-based access control, and secure handling of PII
• Establish tracing standards using OpenTelemetry, LangSmith, or LangFuse to transform the "black box" of agent loops into transparent, debuggable workflows
• Define automated metrics for measuring response quality, reasoning reliability, hallucination rates, and task success rates
• Coach architects, engineers, and technical leads, embedding governance and AI best practices into platform design from day one
• Support sales and solution teams with credible architecture guidance, participating in technical positioning for high-value strategic opportunities

Exadel is an Equal Opportunity Employer committed to inclusion across minority, gender identity, sexual orientation, disability, age, and more. Reasonable accommodations are available to enable individuals with disabilities to perform essential functions. Duties and responsibilities may evolve based on business needs.