Senior AI Platform Engineer – Agentic AI

Qualifications

• YOU MUST HAVE
• 8+ years of software, data platform, ML platform, or AI engineering experience, including designing, building, and supporting production-grade systems.
• Strong Python software engineering expertise, including API/service development, testing, debugging, packaging, design patterns, and modern software development best practices.
• Experience building cloud-native distributed applications and platforms using containers, Kubernetes, public cloud technologies, CI/CD pipelines, observability, and reliability engineering.
• Hands-on experience developing and deploying AI/ML and Generative AI solutions, including model serving, lifecycle management, monitoring, operational support, and production deployments.
• Experience building LLM and agentic AI applications, including RAG/grounding, enterprise data integration and retrieval (structured and unstructured data), orchestration frameworks, and tool/API integration.
• WE VALUE
• Experience with LangGraph, LangChain, Semantic Kernel, AutoGen, or equivalent agent orchestration frameworks.
• Experience designing multi-agent systems, agent tool ecosystems, planning/reasoning workflows, agent memory, or human-in-the-loop patterns.
• Experience with AI evaluation and observability platforms such as LangSmith, MLflow or equivalent tooling.
• Experience with Azure Machine Learning, Azure AI services, Databricks, Kubernetes, or comparable AI/data platforms.
• Experience with vector databases, semantic search, knowledge graphs, RDF/SPARQL, Neo4j, ontology engineering, or GraphRAG-style knowledge systems.
• Experience building AI systems that integrate with industrial IoT, operational technology, building systems, digital twins, or high-volume telemetry.
• Experience building developer platforms, SDKs, internal frameworks, or reusable engineering capabilities used by multiple development teams.
• Bachelor's or advanced degree in Computer Science, Engineering, Artificial Intelligence, or a related discipline, or equivalent practical experience.

ABOUT HONEYWELL

Responsibilities

• Agentic AI Platform
• Design and build reusable enterprise platform capabilities for developing, deploying, and operating AI agents and multi-agent workflows.
• Build frameworks and reusable components that allow agents to securely interact with enterprise APIs, data platforms, applications, tools, and knowledge sources.
• Develop orchestration patterns using agentic AI frameworks, without coupling the platform to a single framework.
• Build services supporting RAG and enterprise knowledge retrieval, including document ingestion, retrieval pipelines, vector/semantic search, grounding, citations, and contextual enrichment.
• Develop systematic approaches for agent and LLM evaluation, including task success, response quality, grounding, tool-use correctness, latency, reliability, and cost.
• Implement observability and tracing for agent execution so developers can understand agent decisions, model calls, tool calls, failures, performance, and resource consumption.
• Design, build, and operate scalable platform services and APIs using Python and FastAPI or comparable production frameworks.
• Build reusable APIs, SDKs, libraries, templates, and reference implementations that make it easier for engineering teams to develop and deploy AI applications.
• Design cloud-native AI platform services using technologies such as Azure, Kubernetes, Databricks, Azure
• Machine Learning, MLflow, or equivalent platforms.
• Develop production capabilities for model and application deployment, configuration, versioning, experimentation, evaluation, and lifecycle management.
• Build automated CI/CD and testing pipelines for AI services, agent workflows, prompts/configurations, and supporting platform components.
• Convert prototypes and notebook-based experimentation into modular, testable, maintainable, production-quality software.
• Design platform patterns for maintaining data quality, lineage, access control, and traceability across AI workflows.
• Build AI services with production software engineering practices covering testing, security, scalability, reliability, and maintainability.
• Implement end-to-end observability, tracing, monitoring, and alerting across agent workflows, models, APIs, retrieval systems, and platform services.
• Establish mechanisms for handling agent failures, tool failures, model errors, timeouts, retries, and degraded dependencies.
• Implement appropriate guardrails, access controls, auditability, and human oversight for enterprise agentic AI workflows.
• Optimize AI workloads for latency, throughput, infrastructure utilization, and model/token cost.
• Diagnose complex production issues spanning AI models, orchestration, data, APIs, infrastructure, and applications.