> Markdown version of [/jobs/ext/4822549-ai-engineer-cx](https://www.wearedevelopers.com/jobs/ext/4822549-ai-engineer-cx). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # AI Engineer - CX - **Company:** Cisco Systems, Inc. - **Location:** San Jose, CA, United States - **Experience:** Expert - **Salary:** $203,000 - $258,600 - **Contract:** Permanent contract - **Skills:** LangGraph Framework, Artificial Intelligence, Amazon Web Services, Microsoft Azure, Routing, Regression Testing, OpenAI, Search Technologies, Software Engineering, Data Streaming, Reinforcement Learning, Retrieval-Augmented Generation, Transfer Learning, Delivery Pipeline, Large Language Models, Multi-Agent Systems, Prompt Engineering, Backend, Agentic-AI, Containerization, AI Platforms, Kubernetes, Low Latency, CrewAI, AutoGen, Ollama, Triton Inference Server, Invoking Functions, Google Gemini, Model Context Protocol, VLLM, Semantic Kernel, Cisco, Docker, Agent2Agent Protocol - **Published:** November 21, 2026 - **Apply:** https://jobs.localjobnetwork.com/apply/add/88922284/1 ## About the Role * Bachelor's with 6+ years of related experience, or Master's degree with 3+ years of related experience. * Experience building and deploying production LLM applications, including tool/function calling, streaming, multi-turn state management, and RAG architectures using vector search. * Design experience of multi-agent architectures (orchestration, routing, shared context, and error recovery) and working with agent integration protocols such as Model Context Protocol (MCP) or Agent2Agent (A2A) standards. * Experience with language model mechanics (transformers, tokenization, context windows, embeddings), model fine-tuning concepts, and comparative evaluation techniques. * Experience developing automated evaluation datasets, scoring methodologies, and regression test suites to systematically benchmark agent accuracy and latency. * Experience working in cross-functional teams, delivering results in fast-paced environments, and effectively communicate technical concepts. Preferred Qualifications * Hands-on experience with modern agent frameworks and orchestration patterns (e.g., LangGraph, AutoGen, CrewAI, Semantic Kernel). * Familiarity deploying and optimizing self-hosted inference runtimes (e.g., vLLM, Ollama, Triton Inference Server) alongside hosted model endpoints (Azure OpenAI, Anthropic, Gemini, AWS Bedrock). * Experience with parameter-efficient fine-tuning (PEFT/LoRA), prompt optimization frameworks, or reinforcement learning from human/AI feedback (RLHF/RLAIF). * Strong proficiency with containerized deployments using Docker, Kubernetes, Helm, and automated CI/CD deployment pipelines. ## Description Cisco's CX AI Incubation and Agents team builds the foundational AI platform and intelligent agent services powering customer experiences across Cisco. Our charter spans the development of AI-native systems-from custom agent orchestration and tool integration to evaluation frameworks and high-scale model serving across cloud, on-premises, and edge environments. We operate as an agile, high-velocity engineering team turning breakthrough AI research into dependable, enterprise-grade software. Your Impact As an AI & Agentic Systems Engineer, you will design, build, and deploy production-grade multi-agent systems and foundational AI services for enterprise network operations, delivery intelligence, and edge automation. You will partner with product, platform, and architecture teams to take open-ended architectural concepts and translate them into resilient, high-throughput agent runtimes. This role combines hands-on backend software engineering with deep applied AI, focusing on real-world reliability, state management, and governed autonomous execution. Key Reponsilbilities: * Build agent services: tool-calling loops, conversation and run state, streaming responses, and human approval for sensitive actions. * Design and implement multi-agent systems, including orchestration, delegation, routing and hand-off between specialized agents. * Build and expose tools and data through the Model Context Protocol (MCP). * Enable agent-to-agent communication through the Agent2Agent (A2A) protocol, including agent discovery, capability description and task exchange. * Integrate multiple model providers (hosted and self-hosted) behind a provider-neutral interface. * Build evaluations that measure agent quality and catch regressions. * Work with product, platform and architecture teams to turn open-ended ideas into agreed designs, then drive them to delivery.