AI Architect

Job Title : AI Architect - Conversational AI.

About the Role:

We are looking to onboard an AI Architect for CX workloads to lead the design and implementation of enterprise-grade conversational, generative, and agentic AI solutions. This role is responsible for architecting scalable, secure, highly available chatbot, voicebot, IVR, and AI agent systems, integrating LLMs and enterprise platforms, and driving intelligent automation initiatives across Contact Center and Customer Experience (CX) environments. The ideal candidate combines deep Conversational AI, Contact Center, Agentic AI, and enterprise architecture expertise with strong system design capabilities and a research-driven mindset to solve complex enterprise challenges. The candidate should have experience taking AI solutions from concept and prototype through production at enterprise scale, including modernization and migration of existing Contact Center ecosystems.

Key Responsibilities

Conversational AI & GenAI Architecture

• Design and architect enterprise conversational systems including chatbots, voicebots, IVR automation, AI agents, and agent-assist solutions.
• Define conversational AI architecture standards including intent and interaction modeling, entity strategy, context management, fallback handling, escalation, human handoff, and multi-channel experiences.
• Lead implementation of GenAI and Agentic AI solutions using LLMs, Retrieval-Augmented Generation (RAG), tool calling, prompt orchestration, memory, multi-agent workflows, and autonomous/semi-autonomous agent patterns.
• Design production-grade agent architectures with appropriate guardrails, permission controls, human-in-the-loop mechanisms, tool governance, and failure-handling strategies.
• Establish prompt engineering best practices, response evaluation frameworks, AI safety guardrails, and continuous evaluation strategies.
• Drive continuous improvement of conversational and agent performance through analysis of historical conversations, customer behavior, operational data, and AI evaluation results.
• Define performance benchmarks for response accuracy, task completion, latency, cost, reliability, and overall customer experience.

Contact Center Transformation & Modernization

• Lead architecture for large-scale Contact Center migrations, modernization, and transformation initiatives involving legacy IVR, telephony, CCaaS, conversational AI, agent-assist, and enterprise backend systems.
• Define migration strategies covering current-state assessment, target architecture, coexistence, phased migration, cutover, rollback, and decommissioning of legacy platforms.
• Architect integrations across CCaaS platforms, telephony/SIP environments, AI platforms, enterprise applications, APIs, and data platforms.
• Evaluate existing Contact Center capabilities and identify opportunities for AI-led automation, self-service, agent augmentation, and operational optimization.
• Ensure solutions support enterprise requirements around scalability, availability, security, compliance, observability, and business continuity.

Agentic AI & Production Engineering

• Architect and implement production-scale Agentic AI systems capable of reasoning, tool execution, workflow orchestration, decision-making, and controlled autonomous actions.
• Define agent architecture patterns including single-agent, multi-agent, orchestrated, event-driven, and human-in-the-loop workflows.
• Establish standards for agent tools, APIs, schemas, authentication, authorization, retries, timeouts, state management, memory, observability, and cost controls.
• Design mechanisms for agent evaluation, traceability, hallucination mitigation, prompt injection protection, policy enforcement, and safe tool execution.
• Drive transition of Agentic AI prototypes and POCs into production-ready enterprise solutions.

System & Solution Design

• Own end-to-end solution architecture (HLD/LLD) for distributed, real-time, conversational, GenAI, and Agentic AI applications.
• Design scalable, event-driven, and API-based integration frameworks with enterprise systems.