Job Summary

We are looking for an experienced AI Solution Architect who bridges client business problems and cutting-edge AI technology — without needing to build models from scratch. You will assess the current technology landscape of our clients, select the right AI models and frameworks, and translate requirements into robust solution architectures. You will own the solution design documentation and work shoulder-to-shoulder with engineering teams to deliver production-grade Gen AI / Agentic AI solutions across operations including order management, billing, collections, Supply chain and contact centre services.

Key Responsibilities

1. Understand Client Landscape

• Engage with client stakeholders to capture current-state processes, pain points, data flows, and technology constraints
• Conduct discovery workshops and process assessments across order management, billing, collections, and contact centre operations
• Map existing systems, integrations, and data sources to identify automation and AI opportunity areas
• Evaluate client readiness: data quality, infrastructure maturity, governance posture, and change capacity

2. Solution Architecture Design

• Design end-to-end Agentic AI solution architectures using existing models (Azure OpenAI, GCP Vertex AI, AWS Bedrock) and frameworks (LangChain/LangGraph, Semantic Kernel, AutoGen, CrewAI)
• Select the right agentic patterns — ReAct, Plan-and-Execute, orchestrator/worker, multi-agent collaboration — for each use case
• Define data ingestion, retrieval (RAG / hybrid search), agent orchestration, tool/function calling, and integration layers
• Conduct build-vs-buy and framework selection analysis; recommend cloud services, vector stores, and third-party tools
• Ensure architecture aligns with client security, data privacy, compliance, and enterprise governance standards

3. Solution Design Documentation

• Produce high-quality design artefacts: solution blueprints, architecture diagrams, agent topology maps, data flow charts, and API/tool contracts
• Author Business Requirement Specifications (BRS), Functional Design Documents (FDD), and Technical Design Documents (TDD)
• Document non-functional requirements: scalability, latency, reliability, cost targets, and responsible AI controls
• Maintain design decisions log and assumption registers; update documentation through delivery lifecycle

4. Delivery Collaboration

• Partner with engineering, data, and DevOps teams to translate solution designs into implementable tasks and sprint backlog

Skill Requirements

Core Architecture Skills

• Strong ability to design and document solution architectures for AI-powered BPO processes — not model building, but expert model and framework selection and integration
• Proficiency with cloud AI platforms: Azure AI Foundry / Azure OpenAI (preferred), GCP Vertex AI, AWS Bedrock
• Hands-on experience with agentic frameworks: LangChain / LangGraph, Semantic Kernel, AutoGen, CrewAI, or LlamaIndex
• Working knowledge of RAG pipelines, vector databases (Azure AI Search, Pinecone, pgvector), and hybrid retrieval strategies
• Familiarity with tool/function calling, MCP (Model Context Protocol), and agent memory strategies

Solution Design & Documentation

• Proven track record authoring TDDs, FDDs, architecture blueprints, and API/integration specs for large-scale programmes
• Ability to create clear, CXO-ready architecture presentations alongside detailed technical documentation for engineering teams
• Experience in structured delivery frameworks: Agile, SAFe, or waterfall-hybrid within consulting or BPO contexts

BPO & Client-Facing Skills

• Experience working directly with clients to assess current-state landscapes and co-design future-state AI solutions
• Understanding of operations: order-to-cash, billing & collections, contact centre, or customer care workflows
• Ability to bridge business stakeholders and technical teams — translating process pain points into actionable AI architectures
• Familiarity with enterprise governance, responsible AI, and data privacy requirements in outsourced environments

Other Requirements