Conversation Architect

We are seeking a Conversation Architect for an initial six-month contract supporting enterprise generative AI and intelligent-agent initiatives. This person will define how AI agents communicate, maintain context, retrieve knowledge, use tools, complete tasks, handle uncertainty, and respond safely across complex customer and employee use cases.

This is not a traditional IVR or call-flow design role. The focus is on generative AI, agent behavior, prompt architecture, knowledge grounding, tool use, evaluation, and scalable conversation standards.

Responsibilities

• Design end-to-end conversational architectures for generative AI agents and virtual assistants.
• Define system prompts, agent instructions, conversation patterns, structured outputs, and behavioral rules.
• Design multi-turn experiences that preserve context, manage state, and guide users through complex tasks.
• Define how agents select and use APIs, tools, workflows, and enterprise systems.
• Design knowledge retrieval and grounding strategies using RAG, search, structured data, and approved content sources.
• Create fallback, clarification, escalation, and human-handoff patterns.
• Establish guardrails for hallucinations, sensitive content, prompt injection, privacy, and inappropriate responses.
• Translate business requirements into conversation architecture, prompt specifications, agent behaviors, and technical requirements.
• Partner with product, engineering, data, design, compliance, and business teams throughout implementation.
• Define evaluation criteria for accuracy, relevance, grounding, task completion, tone, safety, and customer experience.
• Review agent performance, identify failure patterns, and recommend prompt, knowledge, workflow, and product improvements.
• Develop reusable conversation standards, templates, documentation, and governance practices.

Required Qualifications

• Experience designing conversational AI, generative AI, virtual-agent, or LLM-powered experiences.
• Strong prompt engineering skills, including system prompts, reusable prompt patterns, structured outputs, and behavioral constraints.
• Experience designing multi-turn conversations involving intent, context, memory, clarification, and task completion.
• Understanding of API and tool calling, agent workflows, and enterprise-system integrations.
• Experience with RAG, knowledge management, grounding, or retrieval-based AI systems.
• Ability to translate business requirements into clear conversation and technical specifications.
• Strong communication, presentation, facilitation, and stakeholder-management skills.
• Experience collaborating with engineering, product, data, and business teams.

Preferred Qualifications

• Experience defining AI-agent architecture across multiple business domains or use cases.
• Experience with Azure OpenAI, OpenAI, Claude, Gemini, Amazon Bedrock, Google CCAI, or similar platforms.
• Experience creating evaluation rubrics, automated testing, LLM-as-judge workflows, or human-review systems.
• Knowledge of AI safety, prompt-injection testing, hallucination reduction, and responsible AI practices.
• Experience with Python, REST APIs, JSON, JSON Schema, vector databases, or workflow orchestration.
• Experience designing transactional agents that retrieve information, call tools, update systems, or complete business processes.
• Experience establishing prompt libraries, governance standards, quality metrics, and release-readiness criteria.
• Experience in healthcare, financial services, insurance, or another regulated industry.
• Portfolio or work samples demonstrating AI-agent behavior, prompt architecture, conversational workflows, or evaluation strategy.