AI Architect

• Experience in cloud and distributed systems architecture focused on scalability, reliability, observability, and performance.
• designing enterprise AI/ML systems; 1+ years hands-on with GenAI, agentic workflows, RAG, LLM-based integrations, or multi-agent systems.
• Strong expertise with agentic frameworks and tooling (MCP, LangChain, LangGraph,LlamaIndex, autogen, crewai, Agent sdk,OpenAI SDK etc).
• Hands-on experience in modern software development and engineering practices.
• Proven experience integrating APIs and enterprise systems into agentic platforms and workflows.
• Ability to rapidly build AI-driven prototypes, proofs of concept, and demo-ready product experiences.
• Experience defining and governing enterprise architecture standards, patterns, and reference architectures.
• Deep understanding of MCP servers, tool calling, registries, eval pipelines, agent observability, and multi-agent orchestration.
• Hands-on experience with Azure and GCP, including Kubernetes, containerization, identity, networking, CI/CD, and API platforms.
• Familiarity with AIOps/MLOps stacks (MLflow, model registries, vector DBs, semantic layers, feature stores, monitoring).
• Strong knowledge of security, compliance, risk, and Responsible AI (RAI) considerations for enterprise agent systems.
• Demonstrated ability to partner across engineering, data science, product, and security teams to deliver complex AI platform architectures.

• Define and maintain the enterprise reference architecture for agentic platforms (agentic framework, tools, MCP, registries, evals, orchestration, grounding, observability).
• Establish architectural standards and best practices for agent design, tool integration, safety, telemetry, versioning, and lifecycle management.
• Provide architectural leadership for agentic platform engineering teams, ensuring scalability, resiliency, performance, and operability.
• Design and guide integration with semantic layers, embeddings, vector search, knowledge models, and enterprise data products to enable grounded agent behavior.
• Drive architectural direction for low-code/no-code agent-building platforms, ensuring governance, consistency, and ease of adoption.
• Partner with cloud, security, product, and enterprise architecture teams to align agentic platform designs with client's AI governance and RAI principles.
• Define and support agentic SDLC through patterns for evals, safety tests, regression gates, monitoring, and benchmarking.
• Evaluate new agentic frameworks, open-source standards, and orchestration tools to guide build vs buy platform decisions.
• Provide hands-on architectural guidance across engineering and data science teams, enabling scalable, secure, and cost-efficient agent deployment.
• Translate complex business needs into clear technical design patterns and platform capabilities that accelerate agent development.

deepak.kumar@leadstackinc.com